Professional Documents
Culture Documents
edited by
Parasmani Dasgupta
Anthroplogy and Human Genetics Unit, Indian Statistical Institute,
Calcutta, India
and
Roland Hauspie
Laboratory of Anthropogenetics, Free University Brussels, Belgium
v
vi
21 Ethnic and Sex Differences in the Skelic Index among Fijian and
Samoan Children
T. Satake, K. Hattori, E. Kanazawa ................................................... 269
25 Short-term Growth
M. Hermanussen 321
ix
x
One morning in 1969, out of the blue, I received a letter which both distressed and
astonished me. It was from a Prof. S.R. Das in Calcutta, who requested me to
accept, for eventual analysis, a mountain of anthropometric data he had accumulated,
as he was ill and did not expect to survive to analyse it himself. The data provided
the astonishment; twenty-two anthropometric characters recorded every six months
or a year, over a period of 14 years, in a mixed longitudinal study of some 560
children, aged six months to twenty years. Most were in families with siblings also
in the study, and every child was measured every time by S.R. Das himself. The
archive was unique, combining the personal anthropometry of R.H. Whitehouse in
the Harpenden Growth Study and the family approach of the Fels Growth Study.
This was a study of which neither I, nor anyone of my acquaintance, had heard.
Even in India, Prof. Das' work was scarcely known. It turned out Das was a
scholarly man, quiet and unassuming, absolutely committed to his Sarsuna-Barisha
Growth Study,just the obverse of the professional showman.
Clearly this was not a request I could refuse, although I already had in hand
enough projects to occupy Siva himself. So, when the packing cases of data sheets
(not even punched cards) arrived, I put them, not in a storage cupboard, but under
the long table in my small office, where I saw them every time I looked up from my
desk, and where visitors barked their knees on them when they ate their sandwiches.
I felt more and more guilty at their neglect, but reality was reality, the impossible
remained impossible. Finally, a young human biologist from Brussels obtained a
Fellowship to work with me for a year. He brought no particular project of his own,
and the association of Roland Hauspie and Sudhir Ranjan Das began, for in the
meantime Das had recovered, and was working in official retirement, at the Indian
Statistical Institute. He would live to be nearly 90.
Thus history: a story of near-universal neglect of the work of a truly dedicated
man until late in life, a story perhaps more familiar in the Arts than in Sciences.
Now, however, we have this volume dedicated to Professor Das' memory, edited by
Dr. Parasmani Dasgupta, S.R. Das' former postdoctoral student, and Dr. Roland
Hauspie. It contains no less than 27 contributions, covering the whole of Human
Auxology. It is highly international in the provenance of both authors and subjects;
from Kathmandu to Caracas, Oaxaca to Alice Springs. There are papers on the
history of Auxology, on the modelling of the individual growth curve, the
construction of population growth references, growth as a measure of population
well-being, secular trend, and the much neglected subject of the relation between
mental and physical development. And, as the advertisements always say, much
more.
I salute the two editors on the successful completion what was clearly a labour of
love, and I commend this book most warmly to what I hope will be a wide circle of
readers amongst, Anthropologists, Educationists, Human Biologists and Paedia-
tricians around the world.
I.M. Tanner
xiii
EDITORS NOTE
xv
xvi
L.D. voss
Early Bird Research Centre, Derriford Hospital, Plymouth, United Kingdom
Aristotle was interested in the mechanism of growth and offered two possible
reasons why growth might be stunted - violent exercise or intercourse before puberty
(Tanner 1981). Likewise, the tall muscular youths described in the Gallic wars were,
in Caesar's opinion, the result of a chaste adolescence! Around this time, the Chinese
philosopher Wang Ch'ung produced an intriguing explanation for height differences
between individuals. It was, he said, due to variation in vital fluid - only when he
had the full amount could man reach his proper height of ten feet and live to one
hundred years. More often than not, some of this fluid was lacking, resulting in an
average height of only six to seven feet (presumably Chinese feet), and a much
reduced lifespan (Boyd 1981).
In contrast, the Romans, and the Greeks before them, must have had at least
some basic knowledge about factors affecting growth. They would deliberately stunt
the growth of slave children so as to fetch a higher price in the market (Tietze-Conrat
1957). There are no records of the measuring instruments used, but the historian
Suetonius, gives detailed descriptions of the shape and height of several historical
figures (Suetonius trans. Rolfe 1928). Augustus was five feet nine inches (a Roman
measure - a little under five feet seven inches in our terms), yet is judged to have
been 'short of stature'. Indeed, his shoes were 'somewhat high-soled to make him
look taller than he really was'. His short stature, however, was noticeable only by
comparison with some taller person standing beside him' Caesar, on the other hand,
was 'tall of stature' with 'shapely limbs'. There was a persistent belief in those days
that certain races had once been much taller and stronger. Tacitus described the
Germanic tribes of old as tall, red-blond and blue-eyed,· and in his account of the
Gallic war, Caesar still found much to admire in the 'noble savage' existence of the
German youth.
3. THE RENAISSANCE
This period did much for art but little for scientific progress in the study of growth
and development. Scholars rediscovered the ancient Greek texts, and the invention of
printing merely served to spread their largely inaccurate ideas. The preoccupation
with the seven ages of man continued, likewise the four elements: earth, air, fire and
water, thought to be present in all matter. Nevertheless, by Shakespeare's time, it
was known that certain plants had the effect of stunting growth, thus satisfying the
THE MEASUREMENT OF HUMAN GROWTH 5
demand for dwarfs in the royal courts and travelling fairs of Renaissance Europe
(Kelnar 1990). Leonardo da Vinci produced the first accurate drawings of the foetus
and he, together with other artists, began to represent infants and children in a more
sympathetic and realistic manner. In Velasquez' painting of Las Meninas, one can
even distinguish, on the same canvas, a normal child, a hypopituitary dwarf and an
achondroplasic dwarf. Sir Francis Bacon, the 16th century philosopher, also
attempted a more scientific approach to the subject of growth, maintaining that
satisfactory growth required three things: adequate but not excessive nutrition, the
right kind of food (that is, not too dry) and natural heat. This natural heat could be
stimulated by exercise, and therefore 'much Going to Schoole', he wrote, 'where
they sit so much, hindered the Growth of Children' (Tanner 1981).
The real impetus behind the measurement of absolute height, came from the
military, and the need for a powerful fighting force. A contemporary illustration of
recruits being measured for the Duke of Sachsen-Weimar's army in 1779, shows,
interestingly, a remarkably modem looking stadiometer together with faultless
measuring technique (Tanner 1981). Data from recruits allows the modem researcher
both to study secular changes in height, and to look for environmental factors
associated with those changes. Records show clearly that the young, mainly working
class boys, measured on entering the navy from 1770 to 1870 (solely for purposes of
identification, should they desert) were extremely small by today's standards, and
nearly six inches shorter than upper class recruits attending a military college (Floud,
Wachter and Gregory 1990).
As yet, the idea that growth data might be used for the purposes of social reform
had not arisen. The 18th century, known as the century of the Enlightenment, was
characterised, however, by a real attempt to replace mere speculation about the
physical world with observation and measurement, albeit with varying degrees of
scientific rigour. One such case concerned some soldiers who had been discharged for
being too short. A clergymen, curious as to why they had apparently shrunk,
suspected the cause to be diurnal variation in height. The ultimate proof was
presented to a meeting of the Royal Society, ' I tried myself.. and found it in like
manner... at Eleven in the Morning I sat down, and fixed an Iron Pin so as to touch
it.... After that, I fatigued myself for half an Hour with a Garden-Roller, and the
Consequence was, that at 12 Ho. 30 Min. I could not reach the Nail sitting, by
about 5 Tenths of an Inch ... '. He also confessed that he had once measured his horse
before and after riding 20 Miles, but' could not perceive the least difference in her
height.' (Wasse 1724).
plotted, is indistinguishable from any modern day growth chart (Scammon 1927).
Seasonal variations in growth were noted as well as diurnal variation in height. After
one particular all night party, the young man was observed to have shrunk (Boyd
1929). In view of the imprecision inherent in the measurement of height, however
(Voss, et at. 1990, Voss, et at. 1991, Voss and Bailey 1997), one suspects that the
observation was quite fortuitous.
From the 18th century there had been a renewed interest in the classification of races
and attempts were made to substantiate reports of the early explorers by actual
measurements. Considerable ethnic differences in height were reported (Boyd 1981).
Lapps were the shortest, with the average male, allegedly, only four and a half feet,
and the Patagonians the tallest, with the largest men apparently over six and a half
feet tall. By the end of the 19th century, anthropometry had become the tool of all
new schools of physical anthropology, and for a long while, efforts to 'map' the
human body occupied those academics who were interested in the origins and
evolution of man. A more sinister development came when imperialist nations
began to collect anthropometric data in order to rank every race according to
presumed differences in capacities. With some distortion of data regarding brain sizes,
scientists were able to 'prove' what they had always believed in so vehemently,
namely, white supremacy. Paul Broca stands accused of using numbers, not to
generate new theories, but to illustrate a priori conclusions (Gould 1981).
Nineteenth century Europe now turned its attention to the health and welfare of
children, when physicians realised how appalling working and living conditions,
rather than climate, were stunting their growth. Villerme showed that the mortality
rate from all causes was greater in the poor districts of Paris than in the richer parts
of the country, and went so far as to suggest that government policy was directly
responsible for the stature of its people (Villerme 1828). In 1876, Pagliani also
noted that children from poorer homes were smaller and lighter than those from the
higher social classes, and was able to demonstrate, for the first time, that the growth
of destitute children improved after they were taken into care (Bogin 1988). Rapidly
accumulating social data in England was seized upon by political reformers and
opponents of the government, realising that short stature could be used as an index,
not yet of organic disease, but of social disadvantage. In particular, the plight of
young children in factory work was noted. There was, at that time, no comparative
data to demonstrate just how small these children were, but we do know that the
mean height of English working class children then was less than that in most
developing countries today (Tanner 1981).
The 19th century also heralded the start of the eternal nature versus nurture debate,
between those who believe the major determinant of growth and development to be
genetic, and those who consider that it is largely shaped by the environment. Broca,
the French anthropologist was a staunch hereditarian. Whilst he conceded that
poverty and nutrition might have some temporary effect, he claimed that they merely
slowed down the rate of growth, delaying the acquisition of final adult height, which
was fixed by the genes (Tanner 1981). To add to this debate, in 1889, Galton
published Natural Inheritance, his seminal work, in which he sought to demonstrate
the heritability of stature and other physical traits.
The environmentalist cause was taken up by Bowditch, a prominent American
physiologist. The physical superiority of the non-labouring classes seemed, in his
view, to depend more on the greater average comfort in which such children live and
grow up (Boyd 1981). Allowing immigration from southern and eastern Europe was
not, therefore, likely to lead to the gradual physical degeneration of the Anglo-Saxon
race in America, feared by some. To support his argument, Boas showed significant
differences in physical characteristics between adult immigrants who had moved to
the US and their children, born in the new country. The children were always taller
10 CHAPTER 1
and heavier, which he ascribed to better health care and nutrition. The hereditarians
and eugenicists remained unconvinced, however, and in 1921, they scored a major
victory with the passage of the Immigration Restriction Acts (Bogin 1988).
would thus be useful for the regulation of child labour and school entrance ... ' (Boas
1912).
The start of the modem era in terms of medical interest in growth surveillance
can be traced back to Bowditch, a pioneer in the field. In 1875, he started to collect
the heights and weights of many thousands of Boston children. 'It seems probable',
he said, 'that the accurate determination of the normal rate of growth in children will
not only throw light upon the nature of the diseases to which childhood is subject,
but will also guide us in the application of therapeutic measures.' (Bowditch 1881).
The Harvard Growth Study, combined both educatiom.l and medical interests.
Trainee teachers were taught anthropometric techniques and made aware of the
possible effect that illness might have on growth. So persistent was this belief,
among the medical profession too, that in 1934, as many as 61 % of schoolchildren
in New York had their tonsils removed by the age of eleven (Illich 1976).
a great number of measurements.' As Tanner points out, this must make Godin the
only true auxologist! The system of growth surveillance published by Godin was far
from simple, and involved indices of psychological, motor, growth, energy and
developmental power, that could be compared to normative values. Where a child's
index was more than (an apparently arbitrary) seven semesters from the standard for
his age, he was deemed 'abnormal', though Godin did rightly stress that changes in
position were more important than absolute position on a chart.
8. JAMES TANNER
Physical anthropology and clinical paediatrics had once been regarded as separate
disciplines, but it was James Tanner who brought the two together. Very much in
the Galtonian spirit, Tanner felt that one of the chief problems in the field of growth
was to determine 'in what ways do men consistently differ from one another' (Tanner
1953). These differences can be observed or measured, and the monumental Atlas of
Child Growth does both (Tanner and Whitehouse 1982). It contains a wealth of
conventional auxological data, but includes, for the first time, 'photogrammetric
anthropometry', a unique visual record of the growth and development of a large
series of children (healthy and otherwise), thus successfully combining Tanner's
THE MEASUREMENT OF HUMAN GROWTH 13
earlier preoccupation with somatotypes, or body shape, (Tanner 1947, 1964) with
his emerging interest in child health.
The Harpenden Growth Study, from which the 1966 British standards for height
and weight were mostly derived, is familiar to all paediatricians (Tanner, Whitehouse
and Takaishi 1966). It was a remarkable team effort, led by Tanner, though the idea
for the study came originally from a nutritionist in the Ministry of Health. An
orphanage in Harpenden had taken part in a study concerning the provision of food
during war time rationing, but in 1948, the value of continuing to monitor the
growth and development of these children was realised. Tanner, to quote his own
words, was 'the obvious, and indeed, only candidate' for the post (Tanner 1981). A
recent graduate in medicine from London, he was at the time an anatomy lecturer in
Oxford. He had both the credentials and, in particular, the enthusiasm to pursue such
a study, but first took the opportunity, while in the US, to familiarise himself with
the methodological problems of longitudinal studies. On returning to London, he
secured funding for a full-time assistant. A former army officer, Reg Whitehouse,
took up the post, and thus began a partnership that was to last some 28 years.
Whitehouse's contribution to The Harpenden Study was remarkable. As well as
being a meticulous worker, he single handedly measured every child for the entire
duration of the study, that is, from 1949 to 1970. It has been estimated that he took
15 measurements on approximately 9,000 child occasions (Tanner 1981). In spite of
its imperfections, The Harpenden Study remains a landmark study in the history of
growth measurement. Out of it came, not only the first cross-sectional British
Standards for height and weight, in an easily accessible form, but also, for the first
time, Porter's long awaited longitudinal standards that would take account of the
variation in the 'tempo' of growth (Tanner and Whitehouse 1976). The first usable
standards for pubertal development were also derived from the study. These have
never been superseded, and are widely known as Tanner Stages. Not least, the study
produced a new range of anthropometric instruments. Galton had designed one of the
first 'modern' stadiometers - a fixed vertical rule with sliding horizontal headpiece
(Fergus and Rodwell 1874). Whitehouse perfected the design, creating the Harpenden
stadiometer, still very much in use today.
Two key figures this century had a considerable influence on Tanner. The first
was the British scientist, D'Arcy Thompson, whom Tanner recalls having heard
lecture as a schoolboy (Tanner 1981). Biologist, mathematician, philosopher, and by
all accounts an intellectual giant (Bogin 1988), he devoted a lot of thought to the
nature of the growth process and visualised growth as an essentially dynamic process
involving movement through time. Tanner felt that he was the first really to
understand the velocity curve, that is, the continuous nature of growth. The other
great influence on Tanner has been the Swiss paediatrician, Prader. Unlike Tanner, he
is first and foremost a paediatrician, his interest lying in the clinical application of
growth research. Tanner confesses that it was largely due to Prader that he was
persuaded to renew his own interest in clinical matters, eventually becoming
Professor of Child Health and Growth at the Institute of Child Health in London
(Tanner 1981).
14 CHAPTER 1
10. REFERENCES
Bennett, J., Brain, R.. Schaffer, S., Sibum, H.O., and Staley, R., 1994, 1900: The New Age (Whipple
Museum of the History of Science: Cambridge).
Boas, F., 1912, The growth of children. Science, 6,815-818.
Bogin, B., 1988, Patterns of Human Growth (Cambridge University Press: Cambridge).
Boulton, P., 1876, Some anthropometrical observations. British Medical Journal, I, 280-282.
Boulton, P., 1880, On the physical development of children. Lancet, ii, 610-612.
Bowditch, H.P., 1881, The relation between growth and disease. Transaction of the American Medical
Association, 32, 371-377.
Boyd, E., 1929, The experimental error inherent in measuring the growing human body. American
Journal of Physical Anthropology, 13, 389-432.
Douglas, J.W.B., Ross, J.M., and Simpson, H.R., 1965, The relation between height and measured
educational ability in school children of the same social class, family size and stage of sexual
development. Human Biology, 37,178-186.
Encyclopedia Britannica, 1926 (London).
Fergus, W., and Rodwell, G.F., 1874, On a series of measurements for statistical purposes, recently
made at Marlborough College. Journal of the Anthropological Institute, 4, 126-130.
Floud, R., Wachter, K., and Gregory, A., 1990, Height, health and history, nutritional status in the
United Kingdom 1750-1980 (Cambridge University Press: Cambridge).
Goldstein, H., 1971, Factors influencing the height of seven year old children - results from the national
child development study. Human Biology, 43,91-111.
Gould, S.J., 1981, The Mismeasure of Man (Penguin Books Ltd. England).
Gregory, R.L.,1987, The Oxford Companion to the Mind (Oxford University Press: Oxford).
Hacking, 1., 1990, The Taming of Chance (Cambridge University Press: Cambridge).
Hrdlicka, A.,1939, Practical Anthropometry (The Wistar Institute, Philadelphia).
Illich, 1., 1976, Limits to Medicine (Penguin Books Ltd. England).
Kelnar, C.1.N., 1990, Pride and prejudice - stature in perspective. Acta Paediatrica Scandinavia (Suppl)
370,5-15.
Kelvin, W., 1891, Popular Lectures and Addresses, Vol 1. (Macmillan Press) p.80.
THE MEASUREMENT OF HUMAN GROWTH 15
Miller, F.J.W., Knox, E.G., Court, S.D.M., Brandon, S., 1974, The School Years in Newcastle upon
Tyne 1952-1962, Being a further contribution to the study of a thousand families (Oxford
University Press: Oxford).
Porter, W., 1893, On the application to individual school children of the mean values derived from
anthropological measurement by the generalising method. Quarterly Journal of the American
Statistical Association, 1, 576-587.
Porter, W.T., 1923, Percentile charts of the height and weight of Boston schoolchildren. Boston Medical
and Surgical Journal., 188,639-644.
Power, c., 1991, Social and economic background and class inequalities in health among young adults.
Social Science Medicine, 32, 411-417.
Prader, A., Tanner, J.M., von Harnack, G.A ., 1963, Catch-up growth following illness or starvation.
Journal of Pediatrics, 62, 646-659.
Pringle, M.L.K., Butler, N.R, and Davie, R.,1966, 11,000 Seven-year-olds (Longman: London).
Raben, M.S.,1958, Treatment of a pituitary dwarf with human growth hormone. Journal of Clinical
Endocrinology and Metabolism, 18,901-903.
Reading, R., Raybould, S., and Jarvis, S.,1993, Deprivation, low birth weight, and children's height, a
comparison between rural and urban areas. British Medical Journal, 307, 1458-1462.
Rhodes, P.,1985, An Outline History of Medicine (Butterworths: London).
Rona, R.J., Swan, A.V., Altman, D.G., 1978, Social factors and height of primary school children in
England and Scotland. Journal of Epidemiology and Community Health, 32, 147-154.
Scammon, RE., 1927, The first seriatim study of human growth. American Journal of Physical
Anthropology,.10,.329-336.
Suetonius Translated by J.C. Rolfe, 1928, Loeb Classical Library (Heineman: New York) ..
Tanner, J.M., 1947, The morphological level of personality. Proceedings of the Royal Society of
Medicine, XL, 301.
Tanner, J.M., 1953, Growth and constitution. In Anthropology Today, edited by A.L. Kroeber
(University of Chicago Press: Chicaog) p.750-770.
Tanner, J.M., 1964, The Physique of the Olympic Athlete (G. Allen and Unwin Ltd.: London).
Tanner, J.M., 1981, A History of the Study of Human Growth (Cambridge University Press:
Cambridge).
Tanner, J.M., and Whitehouse, R.H., 1976, Clinical longitudinal standards for height, weight, height
velocity and weight velocity and the stages of puberty. Archives of Disease in Childhood, 51, 170-
179.
Tanner, J.M., and Whitehouse, RH., 1982, Atlas of Children'S Growth, Normal Variation and Growth
Disorders (Academic Press: London).
Tanner, J.M., Whitehouse, RH., and Takaishi, M., 1966, Standards from birth to maturity for height,
weight, height velocity and weight velocity, British children 1965. Archives of Disease in
Childhood, 41, 454-471, 613-635.
Tietze-Conrat, E., 1957, Dwarfs and Jesters in Art (Phaidon Press: London Press)
Villerme, L.R., 1828, Memoire sur la mortalite en France dans la classe aisee et dans la classe
indigente. Memoires de I'Academie de Medecine, 1,51-98.
Virey, 1.1., 1816, Geant Dictionnaire des Sciences Medicales, 17,553.
Voss, L.D., 1999, Short but normal. Archives of Disease in Childhood, 81, 370-371.
Voss, L.D., 2000, Growth hormone therapy for the short normal child, who needs it and who wants it?
Journal of Pediatrics, 136, 103-110.
Voss, L.D., and Bailey, B.J.R, 1997, Diurnal variation in stature - is stretching the answer? Archives of
Disease in Childhood, 77, 319-322.
Voss, L.D., and Mulligan, J., 1998, Normal growth in the short normal prepubertal Child. Journal of
Medical Screening, 5, 127-130.
Voss, L.D., Bailey, B.J.R, Cumming, K., Wilkin, T.J., and Betts, P.R., 1990, The reliability of growth
measurement. Archives of Disease in Childhood, 65, 1340-1344.
Voss, L.D., Mulligan, J., and Betts, P.R, 1998, Short stature at school entry - an index of social
deprivation, (The Wessex Growth Study). Child, care, health and development 24,145-156.
Voss, L.D., Wilkin, T.J., Bailey, B.J.R, Betts, P.R., 1991, The reliability of height and height velocity in
the assessment of growth. Archives of Disease in Childhood, 66, 833-837.
Wasse, J., 1724, Concerning the difference in height of a human body, between morning and night.
Philosophical Transactions of the Royal Society of London 33,87-88.
Waterlow, J.C., 1988, Linear Growth Retardation in less developed countries (Raven: New York) pp.
238.
CHAPTER 2
T. SHOHOJIt, T. SUMIYAt
1. INTRODUCTION
Human physical growth is a dynamically changeable and inherently vital
phenomenon. Individual growth has own characteristics. The physical dimension of
growth is widely influenced by heredity and life environment. The fluctuation of age
and size attaining at a special growth phase is large among subjects. Growth patterns
gradually change over time and geography. Thus, an optimal asymptotic growth
model is useful and effective for both data reduction and the characterisation of
individual and average physical growth.
Falkner and Tanner (1986), Johnston et al. (1980), and Goldstein (1979)
describe the methodology of longitudinal and cross-sectional growth studies from
the viewpoint of theoretical and practical problems. Boyd (1980), Tanner (1981),
and Hauspie et al. (1995) cover various approaches on physical growth, and present
fruitful work on the history of human growth study from all over the world and a
large volume of references. Kshirsagar and Smith (1995) illustrate various statistical
techniques for studying growth science and repeatedly measured experiments.
One of the purposes for studying growth models is to establish individual
growth patterns and to predict future individual growth as well as to characterise and
describe physical growth. In selecting suitable growth models, we look for a model
that has as few number of free growth parameters as possible. After estimating
growth curves, we should evaluate and compare their goodness of fit and check the
independence of residuals. Here, we introduce an extended Count-Gompertz growth
model.
Once we choose a suitable mathematical growth model and estimate individual
growth parameters, we can estimate individual biological parameters of the growth
feature. We can compare the means of biological parameters among the samples and,
also, discuss the relationship among biological parameters. It is not clear whether
the relationship among biological parameters may be correlated to time and area or
not.
Little or no correlation exists apparently between adult height and the timing of
the pubertal growth spurt in height. However, the partial correlation between adult
height and age at peak height velocity is negative when height at peak height
17
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 17-32.
© 2001 Kluwer Academic Publishers.
18 CHAPTER 2
velocity is controlled (Qin et al. 1996). We study stable relationships between adult
height and some biological parameters (menarcheal age, age at peak height velocity)
of physical growth in detail.
2. APPROACHES
3. MATERIALS
We use the height and weight of 265 Japanese girls from the Hiroshima Growth
Study sample. Height of these subjects are open to public (Shohoji et ai, 1991).
Most girls were born from 1965 to 1968 and they have roughly one observation in
each spring aged of 6 to 18 years. The age intervals between two successive
measurements are at most 40 months. They all have the records on height and
weight at birth. The maximum and minimum numbers of measurements are 22 and
10, and the average numbers of height and weight are 17.5 and 18.0, respectively.
4. GROWTH MODELS
Let t be the age at examination and let y(t) be the measurement at the age t. We
assume a statistical model such as y(t) = H(t) + et where the random variable et is
normally distributed with the mean 0 and variance al. H(t) is called a growth
(distance) curve. The first derivatives of H(t) with respect to t is called its velocity
curve.
Jenss and Bayley (1937) and Count (1943) propose growth curves during
childhood. Winsor (1932), Marubini et al. (1972) and Marubini and Milani (1986)
GROWTH MODELS AND BIOLOGICAL PARAMETERS 19
A-Bt
compare the Gompertz curve H (t) = e-e and a logistic (autocatalytic) curve
-lIF
generalised logistics curve H(t) = ( 1 + Fe A - Bt ) .
One longitudinal growth model, starting at around 1970, describes an
asymptotic lifetime growth from birth to adulthood by mathematical expression.
Bock et al. (1973) propose a double logistic model, and Bock and Thissen (1980)
suggest a triple logistic growth model from the age of 1
H(t)=a l ! 1- p
1 + e -q(t-CJ)
+1+ p
e -i'2(t- c2)
)+ 1+ a2
e -b.3(t-c3)
. (1)
This can represent the mid-childhood growth spurt. Preece and Baines (1978)
successfully develop a new family of growth models, one of which is
J(t) may satisfy a second order differential equation taking an ageing process into
consideration (Richards 1959, Bertalanffy 1957, Day 1966).
For such a special case that g(t) is a modified Count model and J(t) is the
Gompertz model, we may introduce the Count-Gompertz growth model
dH(t)
h( t) = --;j( ="c
l,(t) + ha(t) (5)
ha(t) = BeA-Bte -e
A-Bt {U - C - Dt - Eln(t + l)}.
160
Growth Curve H(f) 1.4
140 -..
1.2 ~
1.0 E
a
-.. 120
E
()
'-"
..... 0.8 ~
-;=1) 100 .G 0
.~
::c
80 ~ .. :
\
....a.
0.6·g
-
Q)
hc(tr~.. >
t
-- ·Age at MHV·--~:
:
:
:,..!..
,'1 '\,
('I ,
~
~
0.4
-=~
60 ---- AgeatPHV ----~----I~ '~~ .. : ~r. 0.2 0
, , : ha(t) ...... ' 'J ....
------ Age at GC------~--.... ---~-------~... , .... d
.- .- -- ---- - - .- .- - .----- - . ---- .~-!":-.--.t----..- ..- ". -"'-.."-;;..._: -... _~.- - 0.0
40
0 2 4 6 8 iO 12 14 16 18 20
Age (years)
1W o r- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -- - - - .
Growth Curve H(I) 1.4
: Height increment
140
Mid·growth spun Length of : adolescence
: dunng
1.2 .s""
. _~~c;l~~:<:W §
, , 1.0 E
""1 20
E
() E
'-' 0.8 ~
~1()()
Adult Height
o
'g
Of}
.Q3 .t.
e- 0.6
::c: o
80 ~
.:.t. ~
..s
~,
I ...... I ' : \ C)' 0.4
~... ~
-.--- Age at MHV--:I.. --- -----; '--:"' :. h \
(1)\ ...~
W , : , 'i" a
----AgeatPHV-;~;:.---.::F:,'~,'"
, t-$.
"I)
0.2 ~
l'J
.Age at GC -.;:~ ...----~-.-. -.,---
I hs~ ,
r~;':'-"""~~ 0 .0
" I 1 --
- - _ ._ - . . _ _• : o'- _...
_ __ __ __ ._ .- . _ - _ _ lflii . _ _ .. _ _~_.. .
40 '
o 2 4 6 8 10 12 14 16 18 20
Age (years)
The solid increasing function in Figure 1 is the growth curve of height H(t) and
the other solid curve is its growth velocity curve h(t). This velocity is divided into
a childhood component 1Ie(t) and adolescent component ha(t) drawn by the dashed
curves. In Figures 1 and 2, the circle (0) is height and the step function is the
average of growth velocity between successive measurements and (A) indicates her
menarche.
The adult size is an asymptote of a growth curve when the age t tends to
infinity. Let the age at peak height velocity (PRV) and the age at minimum height
velocity (MRV) be, respectively, the age of attaining at the local maximum of
growth velocity curve and the local minimum. We conventionally define the age at
growth completion (GC) as the solution of h(t) = 0.1. Once a growth model is fitted
to individual record, the dimensionality of the record is adjusted and regulated. That
is, the record are projected from a vague measurement space to a growth parameter
space and a biological parameter space. We easily apply usual statistical procedures
to the estimates of parameters.
Some children have a mid-growth spurt. The growth curves (1) and (3) may
represent the mid-growth spurt if there is one, but it might be difficult to estimate a
stable and well fitted growth curve. We now introduce a growth model representing
a mid-growth spurt. Let J(t) and get) be a childhood and a school-child growth
curve, respectively. Let H(t) be the height at the age t in months and let U and
U 0 be the adult size and the maximum height during late childhood school-age,
respectively. We have a growth model such as H(t) = J(t) + Ja(t){ U - J(t)} and
22 CHAPTER 2
H(t) = (1- Ja(t) )(1- Jc(t) )g(t) + U oJc(t)(I- Ja(t») + UJa(t). (6)
When we choose suitable functions for g(t), Ja(t) and Jc(t), we can get an
asymptotic lifetime growth model that may represent the mid-growth spurt if there is
one. We define an extended Count-Gompertz growth model
H(t)
{Aa-Bat)(
={C +Dt + Eln(l + t)~ 1- e-e 1- e-e
Ac-Bct)
5. ESTIMATION
The number of measurements and the ages at examination, in a longitudinal human
growth study, usually vary from subject to subject. It is difficult to identify
individual growth and to compare growth patterns directly by the original
measurements. We can reduce the dimensionality by estimating the growth and
biological parameters with the minimum loss of information on measurements.
We may use the least squares method to estimate growth parameters. We set up
a reasonable convergent criterion for non-linear iteration considering the daily and
seasonal variation of growth increment and accuracy of calculation. A normalised
expected amount of Fisher's information (Rao, 1952) on growth parameters per
observation is introduced which is useful for designing a cohort observation scheme
for non-linear regression (Shohoji, 1982). We can show an asymptotic unbiasedness
by obtaining the biases of estimators.
GROWTH MODELS AND BIOLOGICAL PARAMETERS 23
Table 1. Statistics of growth parameters for height and weight (265 Japanese girls)
160
1.4 '""'
-=
140 1.2 §
J§
1.0 E
EI20
~ ~
0.8e
~100
' dj
'g
0.6 Qj
::r: :>
80
0.4~
o
60 0.2 0
40 ' .. '· ... i·....·..i ~· .... i.. ·.... ·:·.. ·· ..i···.. ··i··· .. ···i·..;;
, , 0.0
o 2 4 6 8 10 12 14 16 18 20
Age (years)
A(61) menarcheal height:O::; 145 cm 8(76) 145 cm < menarcheal height:O::; 150 cm
C(90) 150 cm < menarcheal height:O::; 155 cm D(38) 155 cm < menarcheal height
~ , ~
1.4 .......
.c
50
1.2§
E
~40
~
1.0~
'-'
0.8e
~30 'g
' dj
0.6 -a
~ 20 :>
0.4 ~
10 0.2 0 e
........ ·~ O.O
2 4 6 8 10 12 14 16 18 20
Age (years)
Count-Gompertz model
(MA) 0.521 0.672 0.026 0.325 0.509 - 0.110
(MH) 0.513 0.117 0.701 0.643 0.249 0.275
(AGCH) 0.694 0.121 0.109 - 0.034 0.595 - 0.048
(AH) 0.060 0.720 0.166 0.384 0.184 0.452
(MW) 0.336 0.654 - 0.025 0.386 0.037 0.573
(AGCW) 0.546 0.260 0.643 0.228 0.008 0.446
{AWl - 0.149 0.287 - 0.087 0.469 0.583 0.329
Preece-Baines model
Notes: The correlation coefficients in the upper triangular and the lower triangular are
respectively, for the Count-Gompertz model and the Preece-Baines model.
Let r (menarcheal age, adult height) = 0.026 denote that the correlation
coefficient between menarcheal age and adult height is 0.026 for the Count-Gompertz
model. Let r (menarcheal age, adult height I menarcheal height) = -0.558 represent
their partial correlation coefficients when menarcheal height is held constant (i.e.,
menarcheal height is statistically controlled). And let r (menarcheal age, adult height
I menarcheal height, age at growth completion of height) = -0.845 describe their
partial correlation coefficient when menarcheal height and age at growth completion
are simultaneously held constants (Tables 4 and 5). Also, we have r (menarcheal
height, age at growth completion of height) = 0.117, r (menarcheal height, age at
growth completion of height I menarcheal age) = -0.369 and r (menarcheal height,
age at growth completion of height I menarcheal age, adult height) = -0.798. For
the Preece-Baines model, we have r menarcheal age, adult height) = 0.060, and r
(menarcheal age, adult height I menarcheal height, age at growth completion of
height) = -0.905.
For both growth models, menarcheal age and adult height are highly correlated
with each other in a cross-section of all subjects having the same amount of
menarcheal height and age at growth completion although menarcheal age and adult
height are almost independent as a whole.
GROWTH MODELS AND BIOLOGICAL PARAMETERS 27
Table 4. The correlation coefficients and partial correlation coefficients among the
contra-pairs of biological parameters (265 Japanese girls) derived from the Count-
Gompertz model
Let two growth phases be Fl and F2 and let (al,fl) and (a2,h) be a pair of
the corresponding biological parameters, respectively. For example, we take the ages
for al and a2 and heights for II and h. When the correlation coefficient between
al and h is almost zero or positive and its partial correlation coefficient is negative
in a cross-section of all subjects having the same amounts of the pair (a2,/l), the
special pair of biological parameters (al,h) is called the contra-pair of the
biological parameters (a2, 11) with the negative partial correlation.
From Tables 4 and 5, adult height and menarcheal age are a contra-pair of
menarcheal height and age at growth completion of height. Adult height and age at
peak height velocity are also a contra-pair of height at peak height velocity and age
28 CHAPTER 2
at growth completion of height. A pair of adult height and age at minimum height
velocity is also the contra-pair of height at minimum height velocity and age at
growth completion. Adult height and age interval from peak height velocity to
growth completion are the contra-pair of age at growth completion and height
increment after peak height velocity. These are true for weight growth as well as
height growth and also true for Japanese boys as well as girls.
Table 5. The correlation coefficients and partial correlation coefficients among the
contra-pairs of biological parameters (265 Japanese girls) derived from the Preece-
Baines model
Figure 5 presents the scatter diagram of menarcheal age and adult height and 80%
concentrated ellipsoids by menarcheal height groups. In the legends of Figures 5 and
6, the numbers in the parenthesis between alphabet and symbol show the sample
sizes of the groups. The dotted ellipsoid "A" presents the 80% concentrated
ellipsoid for the menarcheal height group (A: :::;145 cm, 61 girls), menarcheal height
of which is shorter than or equal to 145 cm. Figure 6 shows a scatter diagram of
GROWTH MODELS AND BIOLOGICAL PARAMETERS 29
menarcheal age and adult weight and the 80 % confidence ellipsoids by the
menarcheal weight groups.
170 i C
t;.
165 I +
~
::: 160
-=01)
'Q)
:r:: 155
"3
"0
< 150
? o
+ + .
145 + ...t.. ++.... +
. A .. .+···
9 10 11 12 13 14 15 16
Menarcheal age (years)
Figure 5. Scatterdiagram of menarcheal age and adult height. and the 80% confidence
ellipsoids by menarcheal height groups.
A(61) + menarcheal height:<:; 145 em B(76) 0 145 em < menarcheal height:<:; 150 em
C(90)· 150 em < menarcheal height:<:; 155 em D(38) A 155 em < menarcheal height
75
70
,.....,
~ 65
'-"
~1) 60
.~
~ 55
....
:g 50
<
45
40 L ' ..+0
9 10 11 12 13 14 15 16
Menarcheal age (years)
Figure 6. Scatterdiagram of menarcheal age and adult height. and the 80% confidence
ellipsoids by menarcheal weight groups.
A(34) + menarcheal weight :<:; 35 kg B(82) 0 35 kg < menarcheal weight:<:; 40 kg
C(99) • 40 kg < menarcheal weight :<:; 45 kg D(50) A 45 kg < menarcheal weight
30 CHAPTER 2
Although the correlation coefficient between menarcheal age and adult height is
almost zero apparently for the whole sample, adult height linearly associates with
menarcheal age in a cross section of all girls having the same menarcheal height.
Those who have taller menarcheal height may shift into later menarche and taller
adult than those who have shorter menarcheal height. The taller the menarcheal
height is, the shorter the average length from menarche to growth completion and
the average of growth increment of height during the period are. When we restrict
any menarcheal height groups, those who have younger menarcheal age end up on
average taller adults than those who reach menarcheal age later.
6. DISCUSSION
When selecting a structural growth model, it desires to choose a growth model that
• is evaluated from the viewpoint of goodness of fit (e.g., AIC)
• is interpretable from biological and medical points-of-view
• obtain stable estimates of growth parameters by fitting the model (the asymptotic
variances of the estimators are relatively small).
The sample distributions of estimates of many biological parameters are empirically
normally distributed. The current mixed model technique and other recent
approaches are effectively used for analysing these estimates.
If there exists a mid-growth spurt of growth, the triple logistic (2), JP A2 (3) and
the extended Count-Gompertz (7) may accurately describe the mid-growth spurt.
However, we have to improve the growth models because the coefficient of variation
of the estimates for biological parameters relating to the mid-growth spurt is very
large.
The growth models developed for height growth may be applicable for weight
growth even if the fluctuation of weight is relatively large. Moreover, we need such a
non-linear growth model where height and weight are simultaneously considered for
reducing the number of unknown growth parameters.
When the biological parameters are estimated for some assigned growth models,
there may be statistically significant differences among the means of estimates of the
biological parameters of the target growth models in the case the sample sizes are
large. However, we will get some simple and consistent relationships among
biological parameters for any growth models whenever their growth models are well
fitted within a certain extent.
When studying the relationships among the biological parameters, we have to
consider the true relationship among the biological parameters as well as spurious
relationships. It is, needless to say, regarded as one of essential parts of any growth
model analysis for characterising physical growth from a biological point of view, to
study only true relationship among biological parameters. When we consider the
contra-pair of biological parameters in the case that the relationship between
biological parameters may not be explained directly, it becomes sometime easily to
interpret the relationship of these biological parameters. This contra-pair is very
useful for interpreting physical growth and for predicting the future growth. To study
the relationship among biological parameters is a sort of statistical characterisation of
physical growth from a biological point of view.
Acknowledgement. This work was partially supported by Grant-in-Aid for
Scientific Research (B), Japan Society for the Promotion of Science.
GROWTH MODELS AND BIOLOGICAL PARAMETERS 31
7. REFERENCES
Akaike, H., 1973, Information theory and an extension of the maximum likelihood principle. In the 2nd
International Symposium on Information Theory, edited by 8.N. Petrov and E. Csaki (Budapest:
Akademiai Kiado), p. 267.
Bertalanffy, 1. von, 1957, Quantitative laws in metabolism and growth. The Quarterly Review of
Biology, 32, 217-231.
Bock, R. D., and Thissen, D., 1980, Statistical problems of fitting individual growth curve. In Human
Physical Growth and Maturation: Methodologies and Factors, edited by F.E. Johnston, A.F. Roche,
C. and Susanne (New York: Plenum Press), p. 265.
Bock, R. D., Wainer, H., Petersen, A., Thissen, D., Murray, J. and Roche, A., 1973, A parametrization
for individual human growth curves. Human Biology, 45, 63-80.
Boyd, E., 1980, Origins of the study of human growth. (Portland, Oregon: University of Oregon Health
Sciences Center Foundation).
Count, E. W., 1943, Growth patterns of the human physique: An approach to kinetic anthropometry,
Part I. Human Biology, 15, 1-32.
Day, N. E., 1966, Fitting curves to longitudinal data. Biometrics, 22, 276-291.
Falkner, F., and Tanner, 1. M., 1986, Human Growth: A Comprehensive Treaties. 2nd ed., Volume 3:
Methodology. (New York: Plenum Press).
Goldstein, H., 1979, The design and analysis of longitudinal studies: Their role in the measurements of
change (London: Academic Press).
Hauspie, R. C., 1989, Mathematical models for the study of individual growth patterns. Revue
d'Epidemiologie et de Sante Publique, 37, 461-476.
Hauspie, R., Lindgren, G., and Falkner, F., 1995, Essays on Auxology, (Welwyn Garden City:
Castlemead Publications).
Jenss, R. M., and Bayley, B., 1937, A mathematical method for studying the growth of a child. Human
Biology, 9, 556-563.
Johnston, F. E., Roche, A. F., and Susanne, C., 1980, Human Physical Growth and Maturation:
Methodologies and Factors (New York: Plenum Press).
Jolicoeur, P., Pontier, J., and Abidi, H., 1992, Asymptotic models for the longitudinal growth of human
stature. American Journal of Human Biology, 4, 461-468.
Jolicoeur, P., Pontier, 1., Pernin, M.-O., and Sempe, M., 1988, A lifetime asymptotic growth curve for
human height. Biometrics, 44, 995-1003.
Kshirsagar, A. M., and Smith, W. 8., 1995, Growth curves (New York: Marcel Dekker).
Lindstrom, M. J., and Bates, D. M., 1990, Non linear mixed effects models for repeated measures data,
Biometrics, 46, 673-687.
Marubini, E., and Milani, S., 1986, Approaches to the analysis of longitudinal data, In Human Growth
Vol. 3: Methodology, ecological, genetic and nutritional effects on growth, edited by F. Falkner,
and 1.M. Tanner (New York: Plenum Press) p.79.
Marubini, E., Resele, 1. F., Tanner, J. M., Whitehouse, R. H., 1972, The fit of Gompertz and logistic
curves to longitudinal data during adolescence on height, sitting height and biacromial diameter in
boys and girls of the Harpenden growth study. Human Biology, 44, 511-524.
Preece, M. A., and Baines, M. 1., 1978, A new family of mathematical models describing the human
growth curve. Annals of Human Biology, 5,1-24.
Qin, T., Shohoji, T., and Sumiya, T., 1996, Relationship between adult stature and timing of the pubertal
growth spurt. American Journal of Human Biology, 8,417-426.
Rao, C. R., 1952, Advanced statistical methods in biometric research. (New York: John Wiley & Sons,
Inc), p. 131.
Richards, F. J., 1959, A flexible growth function for empirical use. Journal of Experimental Botany, 10,
290-300.
Shohoji, T., 1982, On the asymptotic properties of the maximum likelihood estimates on a growth curve,
Hiroshima University Statistical Research Group Technical Report, No. 63.
Shohoji, T., Kanefuji, K., Sumiya, T., and Qin, T., 1991, A prediction of individual growth of height
according to an empirical Bayesian approach. Annals of Institute of Statistical Mathematics, 43,
607-619.
Susman, E. P., Murphy, J. R., Zerbe, G. 0., and Jones, R. H., 1998, Using a nonlinear mixed model to
evaluate three models of human stature. Growth, Development and Aging, 62, 161-171.
Tanner, 1. M., 1981, A history of the study of human growth (Cambridge: Cambridge University Press).
32 CHAPTER 2
Winsor, C. P., 1932, The Gompertz curve as a growth curve, Proceeding of the National Academy of
Sciences, U.S.A., 18, 1-8.
CHAPTER 3
l. INTRODUCTION
The methods of drawing references for height, using advanced statistical
methodology based on fitting mathematical models of growth are useful to show
both the pattern and timing of growth in individuals and populations (Ledford and
Cole 1998). There is a set of growth models proposed to fit longitudinal and cross-
sectional growth data of attained height by age which are not based on a set of
parameters. They are often called non-structural models (Bock and Thissen 1980).
The underlying assumption, in models depending on a set of parameters (Count
1943, Jenss and Bayley 1937, Preece and Baines 1978, Shohoji and Sasaki 1987,
Jolicoeur et al. 1988, Deming 1957, Marubini et al. 1971, Thissen and Bock
1990), is that the growth pattern has a basic functional form to which a direct
biological interpretation can be attributed (Berkey 1982, Hauspie, 1989). The
disadvantage of these models is that they sometimes impose a too rigid shape upon
the form of the growth curve (Wember et al. 1992, Zemel and Johnston 1994). This
type of functions are normally used to model individual growth curves rather than to
summarise a whole sample; however, they can be also applied to means (Tanner et
al. 1982) or to selected centiles and, in this way, a set of smooth centi1e curves can
be generated. However, the independent fitting of the centile curves is not a good
solution because some centile lines describe heterogeneous patterns of growth and
the distribution of the centile curves often lacks homogeneity. On the other hand,
fitting non-structural models such as polynomials (Welch 1970) and cubic splines
(Largo et al. 1978), the LMS (Cole 1988, Cole, 1990) and Healy's method (Healy
et al. 1988), give also a good description of the growth process and are good
approaches for estimating centile line curves, spaced appropriately relative to
neighbouring curves. However, this kind of models have the drawback of not being
flexible enough or being uncontrollable at the lower and upper end of the age range.
33
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 33-43.
© 2001 Kluwer Academic Publishers.
34 CHAPTER 3
The LMS-method has been proposed to fit either original data or anthropometric
standards (Cole 1988, Cole and Green 1992) of height and weight and is suitable for
traits which do not necessarily have a monotonously increasing pattern as the BM!
or skinfolds for instance (Cole et al. 1995). The LMS-method performs well in large
samples, but in samples of moderate size, or dealing with reference data collapsed
into few data points, can be difficult to find the median as a steady asymptote, so
some authors (Altman 1993) have proposed an alternative to optimise the
smoothing by applying a function to find the median and then normalise the
standard deviation of height by age by computing the residuals to each data point
and normalising these values by the LMS.
This study was conducted in order to construct cross-sectional growth charts by
applying the LMS-method on the original data of height by age to give a local
reference for Biscay, the western province of the Basque Country. The method was
chosen because it affords suitable smooth centile values of cross-sectional samples of
moderate size. Finally, centiles derived were compared with those obtained from a
mixed-longitudinal sample from the same region (Herruindez et al. 1988).
3. RESULTS
The LMS values obtained after fitting the original data (height by age points) are
shown in Table 1. The PB 1 was used to obtain the standard height by age
SMOOTIllNG CEN'TILES BY TIlE LMS ME1HOD 35
distribution. The fit of the PB 1 on the whole sample divided by sex (Table 2) was
considered successful due to the values of the SEE (0.27 in boys and 0.22 in girls)
and the random distribution of the residuals. In fact, for both sexes the runs test did
not indicate systematic bias in the residuals.
Table 1. Values of the IMS parameters fitted to the original data and to the data after
standardisation with PBl fitted to the central tendency.
Boys Girls
Original data Standardised Original data Standardised
data data
L = -0.104 L = 0.849 L = -0.132 L = 0.904
Age, M S M S M S M S
years
7.5 124.85 0.046 124.26 0.228 124.99 0.051 124.41 0.297
8.5 130.21 0.047 130.55 0.246 130.45 0.051 129.65 0.300
9.5 136.06 0.048 136.59 0.265 135.91 0.050 135.08 0.303
10.5 141.06 0.049 141.21 0.282 141.01 0.048 140.61 0.305
11.5 146.63 0.050 145.99 0.299 146.33 0.047 146.98 0.305
12.5 152.32 0.051 15l.l4 0.314 150.82 0.045 152.55 0.304
13.5 157.43 0.051 156.99 0.321 154.54 0.043 156.75 0.298
14.5 162.27 0.051 163.97 0.321 158.02 0.041 159.77 0.292
15.5 167.23 0.049 170.20 0.315 160.45 0.039 161.21 0.288
16.5 171.29 0.048 172.78 0.307 162.52 0.039 16l. 95 0.285
17.5 175.20 0.047 173.64 0.298 164.19 0.041 162.27 0.285
18.5 178.88 0.047 173.86 0.290 165.84 0.043 162.42 0.283
chi- 1.836 2.757 3.401 6.929
square
Table 2. Values of the function parameters and standard error (ME), obtained by
fitting Preece Baines model I (PBI) to the central tendency of the data.
The LMS values served to generate sets of seven centiles, from the 3rd to the
97th for boys and girls, displayed in Figures 1 and 2. Figure 1 displays the first step
of the fitting of height by age based on the whole sample leading to a bad estimation
of final mean height. Figure 2 displays the growth chart obtained the LMS-method
applied to the residuals of the PB 1 fit to the central tendency.
36 CHAPTER 3
200
P97
Boys P90
P75
180 P50
P25
P10
P3
§ 160
1:
0>
om
I
140
120
100
6 8 10 12 14 16 18 20
Age, years
200
Girls
----
P97
180 ~
P90
r
P75
! 0>
160
~~~i
om
I
140
120
100
6 18 20
Age, years
Figure 10 Growth charts for boys and girls showing centile lines fitted by EMS on
unstandardised height by ageo
SMOO1HING CENTILES BY TIlE LMS MErnOD 37
200
Boys
P97
P90
180 P75
P50
P25
5 160 P10
-
..c::
.Q>
Q)
P3
I
140
120
100
6 8 10 12 14 16 18 20
Age, years
200
Girls
180 ~ P97
P90
1'60 t
P75
~
P50
P25
P10
P3
Q)
I
140
120
100
6 8 10 12 16 18 20
Age, years
Figure 2. Growth charts for boys and girls showing centile lines fitted by IMS on the
residuals of height after fitting PBI to the median.
38 CHAPTER 3
Table 3. Smoothed centiles for Boys, obtained by applying the IMS-method to the
standardised data
Table 4. Smoothed centiles for Girls, obtained by applying the LMS-method to the
standardised data
values without rescaling, the final M was rescaled to show the improvement in the
asymptotic trend.
The LMS goodness of fit was checked by comparing the expected and observed
frequencies between the various centile bands, using the chi-square test. The chi-
square test was not significant in boys and girls whether or not standardised height
was employed to fit the model (Table 1).
The final centile values proposed as height by age references for boys (Table 3)
and girls (Table 4) were compared with those previously obtained from the same
region (Hernandez et al. 1988, Ruiz 1989) by computing the percentage of log
differences between centile values: Log (centile IIcentile 2) x 100, taken as
denominator de previous reference. In general, in both sexes, for all ages and centiles
the values of the present study where higher than the previous references as results
from the positive Log differences, with some exceptions at the beginning of the age
range in girls (from 7.5 to 8.5 years in 3rd and 10th centiles) and at the end of the
age range in boys (from 16.5 to 18.5 years in 3rd, 10th, 25th and from 17.5 to 18.5
years in 50th and at 18.5 in 75th).
4. DISCUSSION
In the last years, some longitudinal (Moreno and Carri6 1993, Bemis 1976), mixed-
longitudinal (Hernandez et al., 1988) and cross-sectional (Sandin 1980, Codina
1983, Bemis et al. 1984, Prado et al. 1985, Rosique 1992, Gonzalez 1997, De la
Puente et al. 1997) studies addressed to develop growth charts of height by age have
been conducted in Spain. However, national references for the whole country have
not been proposed and the country tendency is to develop local references by regions
as is the case of the present study. Previously the growth charts used in Spain were
international standards such as those from Tanner and Whitehouse (1976)
recommended by WHO (1978, 1979). Nowadays, the most popular growth charts
used in Spain are those obtained by Hernandez et al. (1988) by means of a mixed-
longitudinal study based on a sample from Biscay. The discussion whether this
region is sufficiently representative of the whole of Spain, to be used as reference, is
already open because a large amount of immigrant children from almost all of the
different Spanish regions can be found in Madrid. Therefore, Madrid should be more
representative of the Spanish population than Biscay, which gathered immigrants
mainly from three distinct regions of Spain (Galicia, Extramadura and Castilla) and
not from the whole country. In spite of this, while lacking other national designs
Biscaian references are suitable. The present cross-sectional growth study could
improve the references of Hernandez et al. (1988) mainly because the ability of cross-
sectional studies of accounting for a greater amount of variance can be useful to
improve longitudinal standards (Tanner and Davies 1985, Hoey et al. 1987), and
last but not least reason, the possible presence of secular trend in height, due to the
lapse oftime past (more than 10 years) between the two surveys.
In spite that Pan and Goldstein (1997) pointed out the usefulness of LMS to
analyse longitudinal data, originally Cole (1988) proposed this method to be used
in a cross-sectional way. Recently, the LMS-method has been applied by Zoppi et
al. (1996) to construct growth charts from a cross-sectional sample of Verona and
proposed centile values as standards for Italy. Cross-sectional growth charts do not
40 CHAPTER 3
have most of the advantages of longitudinal studies (Tanner and Davies 1985) but
they are used on general health evaluations because the recognised great
sensitiveness of height by age data (Berkey et al. 1993). In the present study the
application of LMS on the residuals of the central tendencies of height by age can be
proposed as a good method to construct growth charts for Biscay. It has the
advantages of fitting by LMS the normalised shape of the curves of the centiles of
height distribution and by PB 1 the central tendencies of height by age. Although,
pure cross-sectional growth charts have to be derived from sufficiently large groups of
children (Sinclair 1989), due to the difficult estimation of true values of the extreme
centiles, the combination of PB 1 and LMS to smooth centiles afforded a good
solution in the present Biscaian sample of moderate size. The generated centile lines
have been successfully representative of normal land-marks (Tables 3 and 4) and
proposed as references for height growth of boys and girls
The growth charts based on unstandardised height (Figure 1) showed a poor
solution for drawing centile lines because the right skewness of original data (Table
1), as L values can show, influenced the departure from the monotonous asymptotic
growth of the 50th percentile in the final ages. A larger sample size could correct this
effect. However, Figure 2, based on standardised height by age showed a better chart
because the combination of LMS and PB 1 can cope with the normalisation, in spite
that the distance among centile lines, across ages, was rather homogeneous in both
charts (Figure 1 and 2). In fact, the easier normalisation of centile lines by non-
structural growth models has widely encouraged researchers for using such methods.
The shape of the centile lines (Figure 2) followed a more homogeneous pattern
than those generated by fitting PB 1 on independent raw centi1es as showed other
studies (Rosique et al. 2000). It is well known that for certain ages height does not
present a normal distribution due to increases in popUlation variability and PB 1
cannot afford a solution for this. LMS coped successfully with the excess of
variability of the present sample around puberty by carrying out the necessary
transformation of height distribution and thus determined the various exact centi1es
(Box and Cox 1964).
The application of the LMS to the unstandardised height yields good
correspondence between observed and expected values as shown by non-sigrrificant
chi-square values in both sexes (see Table 1). Thus, the standardisation of height
does not result in a better modelisation, but only a better normalisation, because
LMS better removes the right skewness of the distribution when height is
standardised (change in L coefficient). L with this second approach is near 1 (almost
a normal distribution), as shown in Table 1.
In Spanish studies, centi1e curve smoothing was done, in the past, mainly by
drawing instruments and sometimes done by applying mathematical models only on
selected centile values (Rebato et al. 1993, Martinez et al. 1994), or by free
estimation of centile distribution by non-structural models (De la Puente et al.
1997). Structural models of growth have also been used in order to describe
regional, utban/rural and other ecological differences in Spain (Rosique and Rebato
1995a) by fitting medians or means of height by age of samples taken from the
literature. Limb dimensions and other body segments have also been studied by
means of fitting the PB1 to mean dimensions by age (Rosique and Rebato 1995b)
providing good estimates of mean adult size and average age at peak velocity as in
other studies (Tanner et al. 1982, Cameron et al. 1982, Dasgupta and Das 1997).
SMOOTHlNG CENTILES BY THE LMS ME1HOD 41
5. REFERENCES
Altman, D.G., 1993, Construction of age-related reference centiles using absolute residuals. Statistics in
Medicine, 12,917-924.
Berkey, C.S., 1982, Comparison of two longitudinal growth Models for preschool children. Biometrics,
38,221-234.
Berkey, C.S., Dockery, D.W., Wang, X, Wypij, D., and Ferris, B., 1993, Longitudinal height velocity
standards for US adolescents. Statistics in Medicine, 12,403-414.
Bemis, C., 1976, Sobre el aumento secular de la estatura en Espana. Trabajos de Antropologia, XVIII,
27-32.
Bemis, C., Ochoa, F., Perez, C., Prado, C., Gomez, A., Sandin, M., Sariflena, V., and Varea, C., 1984,
Crecimiento y Desarrollo en niflos rurales de Segovia. Boletin de la Sociedad Espanola de
Antropologia Biologica, 5,7-18.
42 CHAPTER 3
Bock, R.D., and Thissen, D., 1980, Statistical problems of fitting individual growth curves. In Human
Physical Growth and Maturation: Methodologies and Factors, edited by F.E. Johnston, A.F. Roche
and Susanne (New York: Plenum) pp.265-290.
Box, G.E.P., and Cox, D.R, 1964, An analysis of transformation. Journal of the Royal Statistical Society.
Series B, 26, 211-252.
Cameron, N., Tanner, J.M., and Whitehouse, R.H., 1982, A longitudinal analysis of the growth of limb
segments in adolescence. Annals of Human Biology, 9, 211-220.
Codina, M., 1983, Crecimiento y asimetrias en varones barceloneses de edad escolar (6-16 aiios). Ph.D.
Thesis, Universidad Autonoma de Barcelona. Spain.
Cole, T.J., 1988, Fitting smoothed centile curves to reference data. Journal of the Royal Statistical
Society. Series A, 151,385-418.
Cole, T.J., 1990, The LMS method for constructing normalized growth standards. European Journal of
Clinical Nutrition, 44, 45-60.
Cole, T.J., Freeman, J.V., and Preece, M.A., 1995, Body Mass Index reference curves for the UK.
Archives of Disease in Childhood, 73, 25-29.
Cole, T.J., Green, P.J., 1992, Smoothing reference centile curves: the LMS method and penalized
likelihood. Statistics in Medicine, 13,2359-2367.
Count, E.W., 1943, Growth patterns of human physique: an approach to kinetic anthropometry. Human
Biology, 15, 1-32.
Dasgupta, P., Das, S.R, 1997, A cross-sectional growth study of trunk and limb segments of the Bengali
boys of Calcuta. Annals of Human Biology, 24, 363-369.
De la Puente, M.L., Canela, J., Alvarez, J., Salleras, L., and Vicens-Calvet, E., 1997, Cross-sectional
growth study of the child and adolescent population of Catalonia (Spain). Annals of Human
Biology, 24, 435-452.
Deming, J., 1957, Application of the Gompertz curve to the observed pattern of growth in length of 48
individual boys and girls during the adolescent cycle of growth. Human Biology, 29, 83-122.
Gonzalez, A., 1997, Antropologia del crecimiento en la poblacion escolar de la Villa de Bilbao.
Variacion antropometrica e infuencias ambientales. Ph.D. Thesis, Universidad del Pais Vasco -
Euskal Herriko Unibertsitatea. Spain.
Hauspie, R.C., 1989, ModeJes structurels dans l'analyse de la courbe de croissance. Anthropologie et
Prehistoire, 100, 123-128.
Healy, M., Rasbash, J., and Yang, M., 1988, Distribution-free estimation of age-related centiles. Annals
of Human Biology, 15, 17-22.
Hernandez, M., Castellet, J., Narvaiza, J.L., Rincon, I.M., Ruiz, I., Sanchez, E., Sobradillo, B., and
Zurimendi, A, 1988, Curvas y Tablas de Crecimiento. Instituto de Investigacion sobre Crecimiento
y Desarrollo, Fundacion F. Orbegozo (Madrid: Garsi).
Hoey, H.M.C.V., Tanner, J.M., and Cox, L.A, 1987, Clinical growth standards for Irish children. Acta
Paediatric a Scandinavica, Supplement, 338, 3-31.
Jenss, RM., and Bayley, N., 1937, A mathematical method of studying the growth of a child. Human
Biology, 9, 556-563.
Jolicoeur, P., Pontier, J., Pernin, M.O., and Sempe, M., 1988, A lifetime asymptotic growth curve for
human height. Biometrics, 44, 995-1003.
Largo, R.H., Gasser, T., Prader, A, Stuetzle, W., and Huber, P.J., 1978, Analysis of the adolescent
growth spurt, using smoothing spline functions. Annals of Human Biology, 5, 421-434.
Ledford, A.W., and Cole, T.J., 1998, Mathematical Models of growth in stature throughout childhood.
Annals of Human Biology, 25, 101-115.
Martinez, R, Nielsen, A.H., Prado, C., Acevedo, P., Cuesta, R., and Gomez-Lobo, P., 1994, Aplicacion
del modelo de Preece-Baines-I para los datos transversales del crecimiento de la talla en una
poblacion rural (Cuenca, Espana). In Biologia de las poblaciones humanas: problemas
metodologicos e interpretacion ecologica. Actas del VIII Congreso de la Sociedad Espanola de
Antropologia Biologica, edited by C. Bemis, C. Varea, F. Robles and A. Gonzalez (Madrid:
Ediciones de la Universidad Autonoma de Madrid) pp. 671-681.
Marubini, E., Resele, L.F., and Berghini, G., 1971, A comparative fitting of the Gompertz and logistic
functions to longitudinal height data adolescence in girls. Human Biology, 43, 237-252.
Moreno, P., and Carrie., R., 1993, The height and weight of girls in Barcelona (Spain): Longitudinal
data. Acta Medica Auxologica, 25, 59-66.
Pan, H., and Goldstein, H., 1997, Multi-level models for longitudinal growth norms. Statistics in
Medicine, 16, 2665-2678.
SMOOTHING CENTILES BY 1HE LMS ME1HOD 43
Prado, C., Martinez, R., and Nielsen, A.H., 1985, Panimetros longitudinales y transversales en la
poblacion rural y urbana de Cuenca: Estudio ontogenetico y dimorfismo sexual. Actas del IV
Congreso Espafiol de Antropologia Biologica-1. (Barcelona: Ediciones de la Universidad de
Barcelona) pp. 243-252.
Preece, M.A., and Baines, M.J., 1978, A new family of mathematical Models describing the human
growth curve. Annals of Human Biology, 5, 1-24.
Rebato, E., Gonzalez, A., and Rosique, J., 1993, Application du modele I de Preece-Baines (PBI) it
l'etude de la croissance staturale chez deux populations biscaiennes. Bulletin et Memoires de la
Societe d' Anthropologie de Paris, 5,93-102.
Rosique, J., 1992, Estudio transversal del crecimiento en escolares vizcainos. La variacion
antropometrica como componente de la estructura biologica de la poblacion. Ph.D. Thesis.
Universidad del Pais Vasco - Euskal Herriko Unibertsitatea. Spain
Rosique, J., and Rebato, E., 1995a, Comparative study of statural growth in Spanish populations.
American Journal of Human Biology, 7, 553-564.
Rosique, J., Rebato, E., 1995b, Modelizacion del desarrollo somatico diferencial de una muestra de
chic os y chicas de la costa de Vizcaya. Cuadernos de Seccion. Antropologia-Etnografia, 13, 273-
285.
Rosique, J., Gordon, P., Rebato, E., Gonzalez-Montero de Espinosa, M., Callejo, L., Moreno-Heras, E.,
and Marrodan, M.D., 2000, Aplicacion del modelo I de Preece-Baines al estudio auxologico de
muestras contemporaneas e historicas de la poblacion madrileiia. Investigaciones en Biodiversidad
Humana, edited by T.A. Varela. (Santiago de Compostela: Servicio de Publicaciones e Intercambio
Cientifico, Universidad de Santiago de Compostela).
Ruiz, I., 1989, Estudio longitudinal del crecimiento en la etapa puberal en una muestra de escolares del
Gran Bilbao. Ph. D. Thesis. Universidad del Pais Vasco-Euskal Herriko Unibertsitatea. Spain.
Sandin, M., 1980, Consecuencias biologic as de la emigracion: analisis en poblacion escolar extremeiia.
Ph. D. Thesis. Universidad Autonoma de Madrid. Spain.
Shohoji, T., and Sasaki, H., 1987, Individual growth of Japanese. Growth, 51,452-460.
Sinclair, D., 1989, Human Growth after Birth, Fifth edition (Oxford: Oxford Medical Publications) p.27.
Tanner, J.M., and Davies, P., 1985, Clinical longitudinal standards for height and height velocity for
North American children. Journal of Paediatrics, 107, 317-328.
Tanner, J.M., and Whitehouse, R.H., 1976, Clinical longitudinal standards for height, weight, velocity
and stages of pUberty. Archives of Disease in Childhood, 51, 170.
Tanner, J.M., Hayashi, T., Preece, M.A., and Cameron, N., 1982, Increase in length of leg relative to
trunk in Japanese children and adults from 1957 to 1977: comparison with British and with Japanese
Americans. Annals of Human Biology, 9,411-423.
Thissen, D., and Bock, R.D., 1990, Linear and nonlinear curve fitting. In Statistical Methods in
Longitudinal Research, Vol. 2, Time Series and Categorical Longitudinal Data, edited by A von
Eye (New York: Academic Press) pp. 289-318.
Vercauteren, M., 1984, Evolution seculaire et normes de croissance chez des enfants belges. Bulletin
de la Societe royale beige d' Anthropologie et de Prehistoire, 95, 109-123.
Wachholder, A., and Hauspie, R.C., 1987, Clinical standards fro growth in height of Belgian boys and
girls, aged 2 to 18 years. International Journal of Anthropology, 1,327-338.
Weiner, J.S., and Lourie, J.A., 1981, Practical Human Biology (London: Academic Press).
Welch, Q.B., 1970, Fitting growth and research data. Growth, 34, 293-312.
Wember, T., Goddemeier, T., and Manz, F., 1992, Height growth of adolescent German boys and girls.
Annals of Human Biology, 4, 361-369.
WHO, 1978, A Growth Chart for International Use in Maternal and Child Health Care: Guidelines for
Primary Health Care Personnel (Geneva, Switzerland: World Health Organization).
WHO, 1979, Measurement of Nutritional Impact. WHOIFAP179.1 (Geneva, Switzerland: World Health
Organization).
Zemel, B.S., and Johnston, F.E., 1994, Application of the Preece-Baines growth Model to cross-
sectional data: Problems of validity and interpretation. American Journal of Human Biology, 6, 563-
670.
Zoppi, G., Bressan, F., and Luciano, A., 1996, Height and weight reference charts for children aged 2-
18 years from Verona, Italy. European Journal of Clinical Nutrition, 50, 462-468.
CHAPTER 4
1. INTRODUCTION
Anthropometric references for infants and young children are among the most used
tools for assessing paediatric well-being. At the population level growth references
are useful for predicting general emergencies related to food and nutrition, assessing
the equity of distribution of economic resources within and between communities,
evaluating the suitability of general weaning practices, and for screening and
following at-risk groups. For individuals, growth references are the mainstay to
growth monitoring and promotion, help identify the optimal timing of the
introduction of complementary foods, often are used to assess maternal lactation
performance, are applied in the detection of growth faltering, and help in the
diagnosis of failure to thrive and excessive growth.
Recently, several concerns were raised regarding the adequacy of currently
existing growth references, particularly that developed by the National Center for
Health Statistics (NCHS) and currently supported by the World Health Organisation
(WHO 1995a). The NCHS data for the first two years of life originated from a
single, ethnically homogeneous North American community where most babies
were artificially fed, while recent research shows that babies following current feeding
recommendations show different growth trajectories (WHO 1995b) while presenting
excellent health status. Additional concerns with the NCHS-WHO curves were that
measurements were taken at wide intervals and that curve-fitting procedures were
outdated. In 1995, the WHO Expert Committee on "Physical Status: the use and
interpretation of anthropometry" recommended the replacement of the NCHS-WHO
reference data with a new international growth reference. The main conclusions of the
Expert Committee have been summarised in an earlier publication (de Onis and
Habicht 1996).
45
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth. Development and Maturation, 45-53.
© 2001 Kluwer Academic Publishers.
46 CHAPTER 4
A WHO Working Group was assembled to develop the protocol for the new
study, which is aimed at building a new international growth reference for assessing
the growth status of populations and of individual children. The present chapter
summarises the main features of this new study. The full protocol (WHO 1997) is
available from the authors on request.
The new reference is intended for assessing the growth of singleton children
under five years of age, throughout the world. Its basic assumption is that infants
following current feeding recommendations are growing optimally. The curves may
therefore be considered as prescriptive (or optimal) standards, as opposed to
traditional normative (or descriptive) standards based on samples of children,
regardless of feeding or other behaviours.
A major concern when proposing a reference based on recommended practices is
how such restrictions may affect other characteristics of the sample. The less
common the recommended practices are in the society where the reference data are
being collected, the more it is likely that other, possibly unmeasurable selection
forces will also be operating. If a reference population is highly homogeneous, the
distribution of values will be unacceptably narrow, resulting in cut-off values closer
to the mean than would occur given an appropriately heterogeneous reference
population. Measures to avoid the selectivity problem will be further addressed
below.
The new curves are aimed at both population and individual uses. Population
uses will include providing a reference for comparing groups of children, either in
terms of means (or medians) or of the proportions of children below or above a given
cut-off point. Individual uses include screening for attained size (weight, length, etc)
on a single occasion, and monitoring growth over time.
2. STUDY DESIGN
The WHO Multicentre Growth Reference Study (MGRS) will include a longitudinal
study from birth to 24 months of age and a cross-sectional study of children aged 18
to 71 months. In the longitudinal study, cohorts ofnew-borns will be followed up
for the first two years of their lives, with frequent assessments of feeding practices
and growth. It will also allow the assessment of self-selection bias (see below) and
provide incremental measurements. Children will be identified at birth and visited at
weeks 2, 4, 6, 8 and then monthly until their first birthday. In the second year,
bimonthly measurements will be taken.
After two years of age, growth is considerably more linear than earlier, and
feeding patterns appear to be no longer critical for growth curve construction. A
cross-sectional design will therefore be adopted for children aged 18 to 71 months,
to avoid the time and cost of carrying a longitudinal study in that age range. Using
18 months as the lower age limit for this group will allow an overlap between the
two studies. Although the curves will be built for children aged up to 59 months, it
is necessary to extend data collection to 71 months in order to obtain reliable
estimates of growth at the end of the fifth year oflife (see Sample Size section).
INTERNATIONAL GROWTH REFERENCE FOR YOUNG CHILDREN 47
3. POPULATION SAMPLES
There will be different sets of eligibility criteria for subpopulations and for individual
children.
• Voluntary exclusion. At any time, mothers not willing to comply with the visit
schedule may leave the study. The reasons for the refusal will be recorded
Children who are excluded from the growth-curve set of the study, due to lack of
compliance or serious illness, are continue to be visited as scheduled, so that
information will be available on their growth. Mothers who refuse to continue are
being asked if they would allow one more visit on the next birthday of the child.
These visits will allow to document how the growth of children who remain in the
study may differ from the growth of other children in the target subpopulation.
5. ANTHROPOMETRIC MEASUREMENTS
Anthropometric measurements taken in the longitudinal study include weight,
crown-heel length and head circumference, arm circumference, and triceps and
subscapular skinfold thicknesses. The timing for these measurements is described in
Table 1. More frequent data collection takes place during the first year - when
growth velocity is greater - with a total of 21 visits per child, including the first
visit within the first 24 hours of life.
7. SAMPLE SIZE
The precision of growth chart centiles is determined by several factors, of which the
most important is sample size. Other factors are also relevant, including study
design (cross-sectional versus longitudinal), the timing of measurements, and the
method of curve-fitting. Many criteria can be used to set the sample size, but four
were considered in the present study. These were a) the precision of a given centile
at a particular age, b) the precision of the slope of the median curve over a given age
range, c) the precision of the median curve overall and the influence of data at
particular ages, and d) the precision of the correlation between measurements in the
same subjects at different ages. The latter criterion is relevant for velocity references.
Sample sizes were calculated for each of these four criteria, and it was found that a
sample size of 200 for the longitudinal study and 200 per 3 months for the cross-
sectional study should provide adequate precision. These are the numbers for each
sex, and are to be obtained by combining data from several sites.
A relevant finding of the sample size calculations was that the first few
measurements, particularly birth weight, have high variance, while between 1 and 4
years this is low. In addition, limiting the study to children under five years results
in increased imprecision during the fifth year. To address the imprecision of the
curve at the extremes, birth weight needs to be over-sampled and the upper age limit
raised. A four-times larger sample at birth and an upper limit of 71 full months for
the cross-sectional study will considerably improve the precision of the curve
throughout the whole age range of interest.
To obtain 400 children of both sexes, given six participating sites, 70 children
will have to fulfil all criteria in each centre. The number of infants to be initially
recruited will depend on the proportions expected to remain eligible (i.e., complying
with feeding recommendations, healthy and willing to participate) until the age of
two years. For example, if a completion rate of 30% is envisaged, then 233 infants
are to be recruited at birth (233 x 0.3 = 70). As a working figure for estimating the
logistics of the study, a sample of 300 eligible new-borns per centre was used. This
will fulfil the requirement that the sample size at birth should be at least four times
larger than the group of70 children to be followed up longitudinally for two years.
In the cross-sectional study, as for the longitudinal study, 70 children per age
range are required. As mentioned, to improve the estimates of growth in the fifth
year, it will be necessary to expand the upper age limit of the cross-sectional study
to 71 full months. The lower age limit was set at 18 months to provide some
overlap with the longitudinal study. Since three-month groupings are used, then
1260 children aged 18 to 71 full months are needed in each centre, allowing for
refusals, a sample of 1400 per centre was used. The sampling methodology for the
cross-sectional study is adapted to each site depending on local circumstances (e.g.,
a household survey in the catchment areas of the maternity hospitals included in the
study).
52 CHAPTER 4
8.2 Variability
The Expert Committee expressed concern that selecting children who were "too
similar" (e.g., in feeding practices) might lead to low variability and therefore to
outer percentile lines that are too close to the median (WHO 1995a). This would
result in over-diagnoses of both growth deficits and excess (e.g., obesity) when the
curve is used in a general population. In order to minimise this problem, data will
be combined from different parts of the world, thus allowing the international data
set to include ethnic and environmental variability that is not present in most
existing growth curves. In addition, there will be no anthropometric restrictions on
eligibility: the subpopulations to be included will have a spread of mean birth
weights, and no anthropometric restrictions will be made on individual children,
therefore ensuring substantial variability in anthropometric measurements at birth.
Recent research showed that breast-fed infants growing in different parts of the world
have remarkably similar growth patterns and low variability compared to the current
NCHS-WHO reference (WHO 2000). This fmding confirmed the earlier results of the
Expert Committee (WHO 1995a) suggesting that concern with lack of variability
may not be as important as originally believed.
INTERNATIONAL GROWTH REFERENCE FOR YOUNG CHILDREN 53
9. CONCLUDING REMARKS
The WHO Multicentre Growth Reference Study is an ambitious undertaking. By
prescribing the nature of the sample to the degree proposed by the protocol, the
recommended approach represents a significant departure from approaches used in the
past to construct growth references. This approach has been adopted because of the
stated intent to create a reference that also approximates a standard. Thus the
proposed reference should be representative of a group whose care reasonably
approaches recommended health practices. At the time of this writing data collection,
involving about 10,000 children, is well underway in six countries representing all
the major world regions: Brazil, Ghana, India, Norway, Oman and the USA. It is
envisaged that the new reference will be available in the year 2004.
10. REFERENCES
de Onis, M., and Habicht, J.P., 1996, Anthropometric reference data for international use:
recommendations from a WHO Expert Committee. American Journal Clinical Nutrition, 64, 650-
658.
Kramer, M.S., 1987, Determinants of low birthweight: methodological assessment and meta-analysis.
Bulletin of the World Health Organisation, 65, 663-737.
Mansbach, l.K., Greenbaum, C.W., and Sulkes, J., 1991, Onset and duration of breast feeding among
Israeli mothers: relationship with smoking and type of delivery. Social Science and Medicine, 33,
1391-1397.
Meyer, M.B., 1979, Breast-feeding and smoking. Lancet, 1,975-976.
Vio, F., Salazar, G., and Infante, c., 1991, Smoking during pregnancy and lactation and its effect on
breast milk volume. American Journal Clinical Nutrition, 54, 1011-1016.
WHO, 1995a, Physical status: the use and interpretation of anthropometry. Report of a WHO Expert
Committee. Technical Reports Series No. 854. Geneva, WHO.
WHO, 1995b, Working Group on Infant Growth. An evaluation of infant growth: the use and
interpretation of anthropometry in infants. Bulletin ofthe World Health Organisation, 73, 165-174.
WHO, 1997, Working Group on the Growth Reference Protocol. A growth curve for the 21st Century:
the WHO Multicentre Growth Reference Study. Geneva: WHO Department of Nutrition.
WHO, 2000, Working Group on the Growth Reference Protocol and the WHO Task Force on Methods
for the Natural Regulation of Fertility. Growth patterns of breast-fed infants in seven countries.
Acta Paediatrica, 89, 215-222.
CHAPTER 5
fEscuela Nacional de Antropologia e Historia, Mexico, DF, :f: Michigan State University,
East Lansing, Ml
1. INTRODUCTION
In an earlier report, the growth and skeletal maturity of primary school children 6-14
years of age living under low socio-economic conditions in a peripheral urban
settlement (colonia) in Oaxaca, southern Mexico, was considered (Malina et al.
1976). Skeletal maturity was assessed with the Tanner-Whitehouse II (TW II) 20
bone method (Tanner et at. 1975). Mean height and weight of the children
approximated the 3rd and 10th percentiles of British reference data, but TW II skeletal
ages (SA) indicated only a slight lag or delay. About 60% of the children had skeletal
ages that lagged behind their chronological ages. The seeming discrepancy between
the growth status of the children (3rd and 10th percentiles) and the relatively slight
delay in skeletal maturity suggested, perhaps, a dissociation of growth in body size
and skeletal maturity associated with chronic mild-to-moderate undernutrition. There
is the possibility that the TW II method for the assessment of skeletal maturity may
not be sufficiently sensitive in samples living under conditions of chronic, mild-to-
moderate undernutrition. On the other hand, there may be population differences in
the rate of skeletal maturation which need to be considered in selecting a method for
assessment.
This report considers the application of the Fels method for the assessment of the
skeletal maturity of the hand and wrist in this sample of primary school children in
Oaxaca and compares skeletal ages based on the two methods. Do the two methods
of assessing maturity of the hand and wrist provide similar estimates of skeletal age
(SA) in children living under impoverished conditions? Three specific issues are
considered. First, differences in skeletal age and chronological age (SA minus CA,
SA-CA) based on the Fels method are compared to previous assessments with the
TW II method. Second, the distributions of SA-CA differences are compared. And
third, the growth status of the children plotted by CA, Fels SA and TW II SA,
respectively, are compared to United States reference data.
55
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 55-{j5.
© 2001 Kluwer Academic Publishers.
56 CHAPTER 5
2. METHODS
Height, weight, and hand-wrist radiographs of primary school children in a peripheral
urban settlement of the city of Oaxaca in southern Mexico were studied in 1972
(Malina et al. 1976, see also Malina and Little 1981). The city of Oaxaca de Juarez,
the capital of the state, is located centrally in the Valley of Oaxaca. The population
was approximately 79,000 inhabitants in 1960 (Welte 1973) and increased to about
160,000 in the mid-1970s (Selby and Murphy 1979). With the rapid urban growth,
colonias populares, which are often initially settled by a sqm~tters' invasion on the
city's edges, accounted for a major portion of the area of the city.
Conditions in the Valley of Oaxaca at the time of the study reflected, in part,
those which characterised the State of Oaxaca in the early 1970s - a predominantly
rural state in which most of the population was poor. The economy of the state was
among the poorest in Mexico and ranked last in per capita production in 1970. The
colonias populares have been characterised as ".. vital, young, bustling, and energetic
places where a heterogeneous mix of classes and ethnic groups try to make a decent
life for themselves despite appalling public health conditions, poor nutritional
status, and a lack of those services that make urban living possible and sometimes
pleasant" (Selby and Murphy 1979, p. 8).
Most residents of the colonia were migrants from predominantly rural areas,
although the growth status of children who moved into the colonia did not differ
from those born and raised in the colonia. The community was, on average, poor
socio-economically. The median income of residents coincided with the average for
unskilled and semiskilled labour in the city of Oaxaca (Chance 1971), and living
conditions were marginal in terms of nutrition and public health amenities (Graedon
1976).
Dietary information obtained from a 24-hour recall by sixth grade children in the
colonia indicated a limited diet which was quite similar to that of a rural community
where corn, beans and chiles were the basic foods of the daily diet (pefia Reyes, et al.
1995). Compared to children in the rural community, the urban sample reported a
diet which included slightly greater intakes of protein and bread, in contrast to
tortillas.
The growth and skeletal maturity of 351 primary school children, 181 boys and
170 girls, 6 through 13 years of age, comprise the basis of this report. The data were
collected in 1972. The protocol of the study was approved by authorities in the
colonia both at the community and school levels. The total sample of school
children included a broader age range, 5 to 18 years; however, sample sizes under 6
years and older than 13 years of age were small. Radiographs of the left hand and
wrist were taken with the co-operation of a local radiologist, using a portable X-ray
unit and non-screen film. The focal distance was 76 cm with the tube centred above
the head of the third metacarpal.
All films were initially rated with the TW II 20 bone method (Tanner et al. 1975)
and subsequently with the Fels method (Roche et al. 1988). The methods of skeletal
maturity assessment are similar in principle. Both entail matching the hand-wrist
radiograph to a set of criteria, but differ in criteria for making assessments and
procedures used to construct a scale of skeletal maturity from which skeletal ages are
assigned. In the TW II and Fels methods, the film is matched to specific criteria for
each bone. The TW method uses 20 bones - the radius, ulna, seven carpals, and the
SKELETAL AGE IN MEXICAN CHILDREN 57
metacarpals and phalanges of the first, third and fifth digits of the hand. The Fels
method uses the same 20 bones, and the pisiform and adductor sesamoid of the first
metacarpal. The criteria are based upon shape differentiation of individual bones,
epiphyseal union, and the attainment of adult morphology. The Fels method utilises
ratios between linear measurements of epiphyseal and metaphyseal widths of the
long bones. Criteria for long bones in the TW method include observations of the
widths of the epiphyses and corresponding metaphyses, e.g., the epiphysis is as wide
as the metaphysis or the epiphysis is wider than the metaphysis, but ratios of linear
measurements of epiphyseal and metaphyseal widths are not used.
The methods differ in scoring. The scores for each of the 20 bones rated in the
TW II method are summed to provide a maturity score. The total score can be
converted to a skeletal age (SA), although the maturity score can be used by itself.
The carpal bones and the long bones, by definition, each contribute 50% to the total
score and in turn to the SA. The scoring system in the Fels method statistically
weighs the contributions of specific indicators depending on the sex and age of the
child. For example, epiphyseal union of the radius may occur over several years,
while the appearance of a specific shape of another bone may be present only for a
short period of time. Hence, the radius is given less statistical weight and the other
bone more weight in calculating the SA at this point in time. In addition, the Fels
method provides a standard error of the estimate for the SA.
3. RESULTS
Children were grouped into whole-year age categories, i.e., 6.00 to 6.99, etc. Sample
sizes per age group, and mean chronological ages (CA) and Fels and TW II 20 bone
SAs are summarised in Table 1. Fels SAs differ significantly (p < 0.05 and p <
0.01) from CA in all age groups of boys and girls (except 13 year old girls). In
contrast, TW II SAs do not differ significantly from CA in all age groups of girls
and do not consistently differ from CA across age groups in boys (differences are
significant, p < 0.05, in three of the seven age groups of boys).
Differences between SA and CA, i.e., SA minus CA, for each method of
assessment are illustrated in Figure 1. Within each CA group, Fels SAs lag
consistently behind CA more so than TW II SAs in both sexes. The SA-CA
differences between methods are significant except at 13 years of age in girls (p <
0.05 and p < 0.01). SA-CA differences are reasonably stable for the TW II method in
both boys and girls from 6 to 13 years. On the other hand, SA-CA differences for the
Fels method increase with age from 6-13 years in boys and tend to decrease with age
in girls after 8 years.
The extent of deviations between SA and CA with each method of skeletal
maturity assessment was also considered in the distributions around the expected,
i.e., SA = CA. Children were grouped into five categories. One category was defined
as "on time" or average with SA within ± 1 year of CA (±1 year). Two categories of
late (delayed) maturity, SA more than 1 year but less than 2 years behind CA (> -1
year) and SA more than 2 years behind CA (> -2 years), and two categories of early
(advanced) maturity, SA more than 1 year but less than 2 years in advance of CA (>
+ 1 year) and SA more than 2 years in advance of CA (> + 2 years), were also
defined.
58 CHAPTER 5
Table 1. Sample sizes and descriptive statistics for chronological ages and skeletal ages
(years) based on the Fels (Fels SA) and TW 11 (TW 11 SA) methods in Oaxaca children
i3
Q)
-1
«
0
~ -26
i I 1
Girls • TWII
_.. -€)-.-.
Fels SA-CA
SA-CA
(3- .... -
~.... 01 ". ...(3.........
(3 ,. ·····(3·····---e ·······.e--..... ~
i3 ·1
Q)
«
0
<i:: -2
CI) 6
Figure 1. Skeletal age-chronological age (SA minus CA, SA-CAY differences with the Fels
and TW11 methods in Oaxaca boys (top) and girls (bottom)
SKELETAL AGE IN MEXICAN CHILDREN 59
60 60
Boys ~ Fels
TWII ~ Boys I2l Fels
501- 6·9 years 50 10·13 years • TWII
....
Q)
40 ....
Q)
40
..c ..c
§ 30 § 30
z z
20
Figure 2. Distribution of skeletal age-chronological age (SA-CA) differences with the Fels
and 1WIl methods in Oaxaca boys 6-9 (left) and 10-13 (right) years of age
60r 60
Girls
501- 6·9 years I • I2l Fels
TWII
~ Girls
50 10·13 years
~ 0
•
Fels
TWII
240[
-
.... 40
Q)
..c
§ 30 § 30
z z
20 ~ 20
Figure 3. Distribution of skeletal age-chronological age (SA-CA) differences with the Fels
and 1WIl methods in Oaxaca girls 6-9 (left) and 10-13 (right) years of age
Distributions of SA-CA differences for the Fels and TW II methods in 6-9 and
10-13 year old age groups are illustrated in Figures 2 and 3 for boys and girls,
respectively. Overall, about 37% of boys and 50% of girls have an SA within ±1
year of CA with the Fels method. Corresponding percentages for the TW II method
are 61 % and 68%, respectively. Distributions of TW II SAs are more symmetrical
than those for Fels SAs. More Fels SAs are classified as late or delayed, but among
boys 10-13 years of age, approximately equal numbers have SAs classified as "on
time" (± 1 year), > -1 year and> -2 years. No boys have an SA > +2 years of CA
with the Fels method compared to 4% with the TW II method. About 2% of girls
have an SA > +2 years of CA with the Fels and TW II methods.
60 CHAPTERS
170_
1601-
Boys
/: P10
1SOt ~~
§ 140
E
f:::~.
•
D
t:;.
CA
FelsSA
TWIISA
100 1 I I
6 7 8 9 10 11 12 13 14
Years
Figure 4. Mean height of Oaxaca boys plotted by chronological age (CA), Fels skeletal
age (Fels SA) and TW II skeletal age (TWll SA), respectively, related to selected
percentiles of United States reference data (Kuczmarski et al. 2000)
170
Girls
_ _ _ P50
1601-
P25
150
1 /~~O
E
0 140
E
.~ 130
I
120
/'" ~""""""'-LJo_
•
I ~
CA
D FelsSA
t:;. TWIISA
6 7 8 9 10 11 12 13 14
Years
Figure 5. Mean height of Oaxaca girls plotted by chronological age (CA), Fels skeletal
age (Fels SA) and TW II skeletal age (TWll SA), respectively, related to selected
percentiles of United States reference data (Kuczmarski et al. 2000)
SKELETAL AGE IN MEXICAN CHILDREN 61
60
Boys
P50
50
P25
P10
~ 40 P5
E
Cl
.Qj
~ 30
• CA
D Fels SA
t:, TWII SA
10LI----~------L-----~----~-----L----~~----~----~
6 7 8 9 10 11 12 13 14
Years
Figure 6. Mean weight of Oaxaca boys plotted by chronological age (CA), Fels skeletal
age (Fels SA) and TW II skeletal age (TWII SA), respectively, related to selected
percentiles of United States reference data (Kuczmarski et al. 2000)
50 P50
Girls
40
I P25
P10
P5
Cl
..:.::
30
E
Cl
.Qj
~
10 • CA
D FelsSA
t:, TWII SA
0
5 6 7 8 9 10 11 12 13 14
Years
Figure 7. Mean weight of Oaxaca girls plotted by chronological age (CA), Fels skeletal
age (Fels SA) and TW II skeletal age (TWII SA), respectively, related to selected
percentiles of United States reference data (Kuczmarski et al. 2000)
62 CHAPTER 5
Mean height and weight of the children based on CA, Fels SA and TW II SA are
shown relative to United States reference values (Kuczmarski et at. 20(0) in Figures
4 through 7. With the exception of 6 year old boys, mean height of the Oaxaca
children falls below the 5th percentile of the reference data when plotted by CA
(Figures 4 and 5). When plotted by TW II SAs, mean height moves closer to the 5th
percentile, more so in boys than in girls. When plotted by Fels SAs, mean height of
boys is at or above the 10th percentile, whereas those of girls is above and below the
10th percentile.
Mean weight of the Oaxaca school children, with the exception of the sample of
6 year old boys, fluctuate between the 5th and 10th percentiles of the reference values
when plotted by CA. Among boys, mean weight plotted by TW II SAs is at or close
to the 10th percentile whereas mean weight plotted by Fels SAs is at the 25th
percentile of the reference data (Figure 6). Among girls, mean weight shift only
slightly when plotted by TW II SAs, except in the 12 and 13 year old samples, but
fluctuates between the 10th and 25th percentiles when plotted by Fels SAs (Figure
7).
- - - Ohio boys
2 r: ............... • - ••••••• Ohio girls
_ Oaxaca boys
I!?
I1l
V- .... _
--
----8---- Oaxaca girls
...... _- .... __ ._---- ....... .
--._-.
~
~
(J)
= 0
~ 1J
I-;' ••••• '21- •• -••• 0 ... ··· '21- ---. _.<:r ••••••
«(J)
CI)
-1
Qi
u..
-2
6 7 8 9 10 11 12 13 14
Chronological Age, years
Figure 8. Differences between Fels and TW II skeletal ages (Fels SA minus TW II SA) in
Oaxaca children compared to Ohio children in the Fels Longitudinal Study (Roche et al.
1988)
4. DISCUSSION
Comparisons of Fels and TW II SAs (Fels SA less TW II SA) in the sample of
primarily American White children of the Fels Longitudinal Study (Yellow Springs,
Ohio) showed that Fels SAs are consistently greater than TW II SAs at each
chronological age during childhood and adolescence (Roche et al. 1988). The
differences increased with age to 6 years in boys and to 7-8 years in girls, and then
SKELETAL AGE IN MEXICAN CHILDREN 63
declined. The differences were attributed to slower rates of skeletal maturation in the
standardising sample for the TW II method.
Fels SAs were more delayed relative to CA than TW II SAs in the sample of
Oaxaca children (Figure 1). Fels SAs also had distributions that were generally
skewed towards late or delayed SAs with few early or advanced SAs (Figures 2 and
3). In contrast to observations for the sample of well-off children in the Fels
Longitudinal Study, Fels SAs were consistently less than TW II SAs in each age
group of boys and girls. The contrasting trends in absolute differences between Fels
and TW II SAs (Fels SA minus TW II SA) in the Fels and Oaxaca samples are
shown in Figure 8. It appears that Fels SAs of Oaxaca boys deviate from TW II SAs
with age from 7 to 9 years and then the difference between methods is rather stable.
On the other hand, differences between Fels SAs and TW II SAs in Oaxaca girls tend
to decrease with age from 8 to 13 years.
Studies utilising the Fels method of skeletal maturity assessment of the hand and
wrist are not available for other samples of Mexican children living under conditions
similar to those of the Oaxaca sample. Both methods of maturity assessment were
applied to a sample of 50 youth soccer players, 7 to 17 years, from low to middle
class socio-economic backgrounds in two urban centres in north-central Mexico
(Pefia Reyes et at. 1994). The height and weight of the boys were, on average,
similar to a mixed-longitudinal sample of middle class Mexican boys. Mean SA-CA
differences with the Fels and TW II methods were virtually identical, -0.49 and -0.47
years, respectively, among 12 boys <11.0 years, and +0.27 and +0.30 years,
respectively, among 14 boys 11.0-12.9 years. Thus, in this small sample of active
boys from better-off conditions than the Oaxaca sample, differences between Fels and
TWII assessments of skeletal maturity were, on average, small. However, Fels SAs
were more heterogeneous - 13 (50%) of the boys had SAs within ±1 year of their
CAs, 8 (31 %) had SAs delayed by > 1 year relative to CA, and 5 (19(%) had SAs
advanced by > 1 year relative to CA. Corresponding numbers and percentages for TW
II SAs were 18 (69%),4 (15%), and 4 (15%), respectively. The trends in this small
sample of active boys must be viewed in the context of the selective nature of the
sport. Older boys (> 13.0 years) in this sample of Mexican youth soccer players (n =
24) and in a sample of elite Portuguese soccer players (Malina et at. 2(00) tend to be
advanced in skeletal maturity assessed by both methods.
Two other studies have applied the TW II 20 bone method to rural samples of
Mexican children living under conditions generally similar to those of the Oaxaca
sample, one in the north of the Mezquital Valley in Hidalgo (Cahuich and Rosado
1989) and the other in Northwest Morelos (Cervantes 1989). TW II SAs in the three
samples of Mexican children 7-13 years of age do not, on average, consistently
differ, and, perhaps more importantly in the context of the present analysis, mean
TW II SAs are not significantly delayed relative to CA. Mean height and weight of
the three samples are also similar, i.e., short and light relative to reference data for
well nourished children. This would suggest, perhaps, that skeletal maturation as
assessed by the TW II method is maintained at the expense of growth, which
suggests some degree of stunting. It is also possible that the rate of skeletal
maturation in low socio-economic, mild-to-moderately undernourished Mexican
children is similar to that of the British sample upon which the TW II method was
standardised.
64 CHAPTER 5
In the earlier analysis of the Oaxaca sample (Malina et al. 1976), it was
hypothesised that nutritional stress during infancy and early childhood may have a
more severe effect on growth in size but may have a minimal influence on skeletal
maturation when ossification centres of the hand-wrist are ossifying, i.e. "appearing"
on a hand-wrist radiograph. However, there is minimal information available for the
assessment of skeletal maturity of the hand-wrist during the first 3 or 4 years of life
when the impact of chronic, compromised nutrition is most profound. The knee, in
contrast, may be the more appropriate area for the assessment of skeletal maturity
early in life, i.e., the pre-school years (Roche, Wainer and Thissen, 1975). Data on
the skeletal maturity of the knee in young children reared under conditions of chronic
undernutrition are not available.
In the context of public health, it has been suggested that " ... comparisons of
skeletal maturity between samples of populations reveals degrees of environmental
disadvantage perhaps more finely than growth velocities ... " (Tanneret at. 1983). If
this suggestion is accepted, one can inquire why, in the case of children raised under
unfavourable conditions, is skeletal maturity assessed by the TW II method not
affected to the same extent as their growth in height? On the other hand, trends in the
Fels SAs of the Oaxaca sample suggest that skeletal maturity is in fact affected by
adverse nutritional and general health conditions during childhood. This is especially
clear in comparisons of the height of the Oaxaca sample plotted by CA, Fels SA and
TW II SA relative to United States reference data, which indicate that their small
body size is consistent with delayed Fels SAs (Figures 4 and 6). In contrast, the
relatively slight delay in TW II SAs among Oaxaca children is not commensurate
with their small body size.
In summary, the two methods for assessing skeletal maturity of the hand and
wrist, the Fels and TW II methods, do not provide similar estimates of SA in this
sample of Mexican children living under generally impoverished health and
nutritional circumstances. Fels SAs are significantly delayed relative to TW II SAs
and relative to CA, and the delay or lateness in Fels SAs is consistent with the small
body size of the children. Application of the Fels method for the assessment of
skeletal maturity to other samples living under similar conditions is needed to
confirm the observations in the present analysis.
5. REFERENCES
Cahuich, M., and Rosado, E., 1989 Los habitos alimenticios en una comunidad rural del Valle del
Mezquital. Tesis, Escuela Nacional de Antropologia e Historia, Mexico, OF.
Cervantes, M.C., 1989 Estudio de crecimiento y maduracion esqueletica en una comunidad rural. Tesis,
Escuela Nacional de Anthropologia e Historia, Mexico, OF.
Chance, J.K., 1971 Kinship and urban residence: Household and family organization in a suburb of
Oaxaca, Mexico. Journal of the Steward Anthropological Society, 2 (no. 2), 2 pp. (mimeo).
Graedon, T.L.F., 1976 Health and Nutritional Status in an Urban Community of Southern Mexico.
Unpublished Doctoral Dissertation, University of Michigan, Ann Arbor.
Kuczmarski, R.I., Ogden, c.L., Grurnmer-Strawn, L.M., Flegal, K.M., Guo, S.S., Wei, R., Mei, Z.,
Curtin, L.R., Roche, A.F., and Johnson, C.L., 2000 CDC growth charts: United States. Advance
SKELETAL AGE IN MEXICAN CHILDREN 65
Data from Vital and Health Statistics, no 314 (Hyattsville, MD: National Center for Health
Statistics).
Malina, R.M., Himes, J.H., and Stepick, C.D. 1976 Skeletal maturity of the hand and wrist in Oaxaca
school children. Annals of Human Biology, 3, 211-219.
Malina, R.M., and Little, B.B., 1981 Comparison of TWI and TW2 skeletal age differences in
American Black and White and in Mexican children 6-13 years of age. Annals of Human Biology,
8,543-548.
Malina, R.M., Pena Reyes, M.E., Eisenmann, J.C., Horta, L., Rodrigues, J., and Miller, R., 2000 Height,
mass and skeletal maturity of elite Portuguese soccer players ages 11-16 years. Journal of Sport
Sciences, 18, 685-693
Pena Reyes, M.E., Cardenas-Barahona, E., and Malina, R.M., 1994 Growth, physique and skeletal
maturation of soccer players 7-17 years. Humanbiologia Budapestinensis, 25, 453-458.
Pena Reyes, M.E., Malina, R.M., Little, B.B., and Buschang, P.H., 1995 Consumo de alimentos en una
comunidad rural Zapoteca en el Valle de Oaxaca. In Estudios de Antropologia Biologia, Volumen
5, R.M. Ramos Rodriguez and S. Lopez Alonso, eds. (Mexico, DF: Instituto Nacional de
Antropologia e Historia), pp. 407-414.
Roche, A.F., Chumlea, W.e., and Thissen, D., 1988 Assessing the Skeletal Maturity of the Hand-Wrist:
Fels Method (Springfield, IL: CC Thomas).
Roche, A.F., Wainer, H., and Thissen, D., 1975. Skeletal Maturity: The Knee Joint as a Biological
Indicator (New York: Plenum).
Selby, H.A., and Murphy, A.D., 1979 The City of Oaxaca. Final Technical Report, Office of Urban
Development, Technical Assistance Bureau (Washington, DC: Agency for International
Development).
Tanner, J.M., Whitehouse, R.H., Marshall, W.A., Healy, MJ.R., and Goldstein, H., 1975 Assessment of
Skeletal Maturity and Prediction of Adult Height (New York: Academic Press).
Tanner, J.M., Whitehouse, R.H., Cameron, N., Marshall, W.A., Healy, M.J.R., and Goldstein, H., 1983
Assessment of Skeletal Maturity and Prediction of Adult Height, 2nd edition (New York: Academic
Press).
Welte, e.R., 1973 Population of Selected Localidades in the Valley of Oaxaca: Censuses of 1960 and
1970 (Oaxaca de Juarez, Mexico: Oficina de Estudio de Humanidad del Valle de Oaxaca).
CHAPTER 6
fPaediatric Clinic, "GB. Rossi Hospital", Verona University, Verona, Italy, :t Obesity
Research Centre, St. Luke'slRoosevelt Hospital, Columbia University, College of
Physicians and Surgeons, New York, USA
1. INTRODUCTION
Quantifying the main body components is integral to the study of growth, as the
assessment of human physical characteristics is important both in the
anthropological and medical fields (Forbes 1962). The human body consists of over
thirty components, sometimes also referred to as compartments, distributed across
four main organisational levels: atomic, molecular, cellular and tissue system (Wang
et al. 1992). The sum of all components at each level of body composition is
equivalent to total body mass (Wang et al. 1995). Body composition is influenced by
endogenous and environmental factors, and is a valid indicator of pathological
changes especially when diagnosing obesity and other nutritional disorders (Le.:
bulimia and anorexia) in humans (Frisancho 1984). Studies of body composition and
various risk factors in growing children should have important implications for
preventive medicine (Forbes 1987 ).
Why is there such concern with paediatric obesity not alone with adulthood
obesity? The reason why may be that the prevalence of paediatric obesity is
increasing in the United States (Kucsmarski 1993, Troiano and Flegal 1998) and
European countries (Crepaldi et al. 1991, Organon et at. 1988). Moreover, even in
developing and newly industrialised countries, the increasing prevalence of childhood
obesity is pronounced (Rossner 1998, Mo-Suwan and Greater 1996). This rising
obesity prevalence is a concern given the health risks associated with paediatric
obesity, including elevated blood pressure, glucose intolerance, hyperinsulinemia,
dyslipidemias and cardiac disease (Faith et at. 1996, Rocchini 1993, Williams et at.
1992). Accurate assessment of body composition is important in many areas of
obesity and nutrition-related research.
The aim of this chapter is to describe and discuss methods for paediatric body
composition assessment. This chapter is broadly divided into three sections. The first
section discusses several cutting-edge methods that are primarily available to research
centres. The second section reviews low cost, practical methods for estimating body
composition. Specifically, we will focus on three methods that can be used in
67
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 67-75.
© 2001 Kluwer Academic Publishers.
68 CHAPTER 6
Recent advances in techniques for measuring body composition have provided DXA
for assessment of whole-body as well as regional measurements of bone mass, lean
mass and fat mass (Mazess et al. 1992). For a more in-depth review of the physical
concept of DXA, see Pietrobelli et al. (1996).
The DXA method evolved from the simpler single-photon absorptiometry
method used to quantify bone mass in areas with little overlying soft tissue
(Cameron and Sorensen 1963). When soft tissue is minimal, photons, emitted from
a radioactive source, are attenuated almost exclusively by higher atomic weight
elements such as calcium and phosphorus in bone.
In 1981, Peppler and Mazess (1981) first introduced the concept of dual-photon
absorptiometry. Two photon energies were passed through tissue and then
attenuation could be measured. This technique has the potential to substantially
improve the feasibility of body composition analysis with children for whom other
laboratory techniques may be impractical (e.g., underwater weighing). Briefly DXA
measurement is based on the differential attenuation of two photon beams as the
various tissues of the body absorb them. DXA requires minimal co-operation from
the participant and is relatively quick (20 minutes for a child). Radiation dosages are
very low « 1 mrem) (Pietrobelli et al. 1998).
The greatest advantage of DXA may be the ability to assess regional body
composition (i.e., trunk, arms, and legs). Nutritional status of diseased individuals
can be evaluated by analysing the individual compartments of the body. To date,
several studies were performed using DXA in paediatric samples and the results have
shown encouraging evidence that DXA is an accurate method of assessing body
composition also in paediatric sample (Goran et al. 1995). Costs vary and are higher
than those for simpler methods such as BIA or anthropometry. DXA is a stationary
instrument suitable for laboratory research and at present its clinical use for soft
tissue analysis is limited (Pietrobelli et al. 1998). However, DXA offers a new
method for the study of skeletal maturation, mineral homeostasis, environmental,
and nutritional factors involved in development and growth (Lapillonne et al. 1997)
because it measures the three body compartments (i.e. bone mineral, fat mass and
lean soft tissue).
reconstruct various tissue volumes including total, subcutaneous and visceral adipose
tissue, skeletal muscle, brain, organs, skin, and bone. Imaging techniques are
expected to provide new insights into the physiology of intra abdominal adipose
tissue and its relation to health (Goran et at. 1995).
These techniques are widely used in research studies to evaluate lipid, hormonal
and other relations to body composition (Heymsfield et at. 1997). CT and MRI have
the ability to quantify 3-dimensional tissue volume. The images are sharp enough
for tissue boundaries to become clear (Sjostrom 1991). None of other currently
available methods can assess tissue-system level body composition components with
the same accuracy as CT and MRI. Both CT and MRI are costly tests and their
application to low-budget and large-scale epidemiological studies is limited.
BMI=W/H2
This index has the advantage of being inexpensive, safe, and easy to obtain. It
only requires a standard scale and an accurate way to measure height. A home
bathroom scale can provide a reliable measurement of body weight. Recently we
compared BMI estimates of body fat against body fat measured by DXA in a large
paediatric sample and found a significant association between the two methods
(Pietrobelli et al. 1998). In addition, BMI is reported in many epidemiological data
sets, making it commonly available for population-level analyses (Harlan 1993). At
the same time, a limitation of BMI is that this index cannot differentiate fat mass
from fat-free mass (i.e., lean tissue) (Lohman 1992, Ellis et al. 1999).
Weight and height reference charts are important for monitoring height and excess
weight, but do not assess nutritional status in children (Rolland-Cachera 1995).
During growth period weight increases with both age and height, and the weight
changes are mainly due to changes in stature rather than those in fatness. However,
changes in BMI reflect changes in body fatness (Rolland-Cachera 1995).
According to previous studies (Himes and Dietz 1994, Harlan 1993, Dietz and
Bellizzi 1999), body weight and height and subsequently BMI are essential
parameters to assess growth and development in children. It is suggested by some
that these parameters be measured regularly from birth to better detect deviations
from normal development (i.e. growth retardation or overweight / obesity) (Forbes
1987).
In 1997 the International Obesity Task Force (lOTF) convened a Workshop on
childhood obesity to explore the strengths and limitations of existing approaches to
the measurement of childhood obesity. The workshop concluded that BMI is a
70 CHAPTER 6
reasonable and, for clinical purposes, preferred measure of fatness in children and
adolescents. The standards used to identify overweight and obesity in children and
adolescents should agree with the standards used to identify Grade I and 2 overweight
in adults (BMI of 25 or 30, respectively) (Dietz and Bellizzi 1999).
• Take on the right side of the body with the child supine on a non-conductive
surface.
• Clean the skin at the electrodes sites.
• Place the sensor electrodes on the dorsal side of the wrist and ankle.
• Place the source electrodes at the base of the second third metacarpaVmetatarsal-
phalangeal joint of the hand and foot.
• The arms and the legs of the child will be abducted approximately 45° to each
other.
Table 4. Features that can be taken into consideration for appropriate choice of a body
composition method.
4. CONCLUSION
There is no single body composition method, which is universally the "best" for
paediatric samples; rather, the clinicians or researcher must weigh the practical
considerations of their assessment needs with the limitations of the methods.
Regardless of which instrument is chosen to assess body composition, it is crucial
for the clinician to meticulously follow the standard guidelines (Dietz and Bellizzi
1999) and protocol associated with each method to limit measurement error. In
addition the conversion formulas and prediction equations selected must be restricted
to the populations from which they were derived and validated (Wagner and Heyward
1999).
The clinician should consider the various factors that can influence body
composition. It is important to know nutritional factors, energy intake, and
composition of the diet, nutrition and hormonal status, food preferences and
behaviour, and the influence of non-nutritional factors. When these are taken all
together with an accurate and precise body composition assessment, it may be
possible to have the possibility to control growth process and to predict adult status
in order to reduce the risk factors of various diseases. We have reported in Table 4
some features that can be taken into consideration for appropriate choice of body
composition method.
5. REFERENCES
Cameron, J.R., and Sorensen, J., 1963, Measurement of bone mineral in vivo. Science, 42, 230-232.
Crepaldi, C., Belfiore, F., Bosello, 0., Caviezle, F., Contaldo, F., Enzi, G., and Melchionda, N., 1991,
Consensus Conference ltaliana: sovrappeso, obesita' e salute (CCI S.O.S.). Annali ltaliani
Medicina Interna, 6, 359-364.
Deurenberg, P., Kusters, C.S.L., and Smith, H.E. 1990, Assessment of body composition by bioelectrical
impedance in children and young adults in strongly age-dependent, European Journal of Clinical
Nutrition, 44, 261-268.
Deurenberg, P., Van der Kooy, K., Paling, A., and Withagen, P., 1989, Assessment of body composition
in 8-11 year old children by bioelectrical impedance. European Journal of Clinical Nutrition, 43,
623-629.
Dezenberg, C.V., Nagy, T.R., Gower, B.A., Johnson, R., and Goran, M.I., 1999, Predicting body
composition from anthropometry in pre-adolescent children. International Journal of Obesity, 23,
253-259.
Dietz, W.H., and Bellizzi, M.e., 1999, Assessment of childhood and adolescent obesity. American
Journal of Clinical Nutrition, 70 (Suppl), 117-175.
74 CHAPTER 6
Ellis, K.J., Abrams, S.A., and Wong, W.W., 1999, Monitoring childhood obesity: assessment of the
weightlheight2 index. American Journal of Epidemiology, 150,939-946.
Faith, M.S., Pietrobelli, A., Allison, D.B., and Heymsfield, S.B., 1996, Prevention of paediatric obesity:
examining the issues and forecasting research directions. In Preventive Nutrition edited by A.
Bednich and R Deckelbaum (Totowa, (NJ), Humana Press) p. 471-486.
Fomon, S.J., Haschke, F., Ziegler, E.E., and Nelson, S.E., 1982, Body composition ofreference children
from birth to age 10 years. American Journal of Clinical Nutrition, 35, 1169-1175.
Forbes, G.B., 1962, Methods for determining composition of the human body. Pediatrics, 29: 477-94.
Forbes, G.B., 1987, Body composition in infancy, childhood and adolescence. In Human body.
composition: Growth, Aging, and Activity edited by G.B. Forbes (New York: Springer Verlag),
125-196.
Foster, M.A., Hutchinson, J.M.S., Mallard, J.R, and Fuller, M., 1984, Nuclear magnetic resonance pulse
sequence and discrimination of high- and low-fat tissue, Magnetic Resonance Imaging, 2, 187-192.
Frisancho, A.R, 1984, New standard of weight and body composition by frame size and height for
assessment of nutritional status of adults and the elderly. American Journal of Clinical Nutrition, 40,
808-819.
Goran, M.I., 1998, Measurement issue related to studies of childhood obesity: assessment of body
composition, body fat distribution, physical activity, and food intake, Pediatrics, 101,505-518.
Goran, M.I., Figueroa, R, McGloin, A., Nguyen, V., Treuth, M.S., and Nagy, T.R., 1995, Obesity in
children: recent advances in energy metabolism and body composition. Obesity Research, 3, 277-
289.
Harlan, W.R, 1993, Epidemiology of childhood obesity. Annals of the New York Academy of Science,
699, 1-5.
Harsha, D.W., Frerichs, RR, and Berenson, G.S., 1978, Densitometry and anthropometry of black and
white children. Human Biology, 50, 261-280.
Heymsfield, S.B., Nunez, C., and Pietrobelli, A., 1997, Bioimpedance analysis: what are the next steps?
Nutritional in Clinical Practice, 12, 201-203.
Heymsfield, S.B., Ross, R., Wang, Z.M., and Frager, 0.,1997, Imaging techniques of body composition:
advantages of measurement and new uses. In Emerging technologies for nutrition research by S.J.
Carlson-Newberry and RB. Costello (New York: National Academic Press), 127-150.
Himes J.H., and Dietz, W.H., 1994, Guidelines for overweight in adolescent preventive services:
recommendations from an expert committee. American Journal of Clinical Nutrition, 59, 307-316.
Houtkooper, L.B., Going, S.B., Lohman, T.G., Roche, A.F., and Van Loon, M., 1992, Bioelectrical
impedance estimation of fat free body mass in children and youth: a cross validation study. Journal
of Applied Physiology, 72, 366-372.
Janz, K.F., Nielsen, D.H., Cassady, S.L., Cook, J.S., Wu, Y., and Hansen, J.R, 1993, Cross-validation of
the Slaughter skinfold equations for children and adolescents. Medicine and Science in Sports and
Exercise, 25, 10780-1076.
Kucsmarski, R.J., 1993, Trends in body composition for infants and children in U.S. Clinical Review
Food Science Nutrition, 33, 375-387.
Lapillonne, A., Braillon, P.M., Delmas, P.O., and Salle, RL., 1997, Dual energy x-ray absorptiometry in
early life. Hormone Research, 48, 43-49.
Lohman, T.G., 1992, Advances in body composition assessment. Current Issues in Exercise Science
(#3), Human Kinetics Press, Champaign, IL
Lohman, T.G., Roche, A.F., and Martorell, R, 1988, Anthropometric standardization reference manual.
Champaign, IL, Human Kinetics.
Lukaski, H.C., 1992, Body composition using impedance methods. In Obesity edited by P. Bjontorp and
B.N. Brodoff (Philadelphia: J.B. Lippincott Company), 67-79.
Mazess, RB., Chesnut, C.H., McClug, M., and Genant, H., 1992, Enhanced precision with dual-energy
X-ray absorptiometry. Calcification Tissue, 33, 353-359.
Mo-Suwan, L., and Greater, A.F., 1996, Risk factors for childhood obesity in a transitional society in
Thailand. International Journal of Obesity, 20, 697-703.
Organon, R.G., Tubol, LB., Zhukovskii, G.S., Perova, N.V., Ilchenko, LN., 1988, Epidemiological
characteristics of dyslipoproteinaemia and certain other risk factors of atherosclerosis and
aeschaemic heart disease in 11- and 14-years children in different climatogeographic zones:
results of a cooperative study. Cor Vasa, 30, 248-256.
Peppler, W.W., and Mazess, R.B., 1981, Total body bone mineral and lean body mass by dual-photon
absorptiometry. I. Theory and measurements procedure. Calcif Tissue, 33, 353-59.
Pietrobelli, A., Faith, M.S., Allison, D.B., Gallagher, D., Chiumello, G., and Heymsfield, S.B., 1998,
Body mass index as a measure of adiposity among children and adolescent: a validation study.
Journal of Paediatric, 132,204-210.
PietrobeJli, A., Formica, C., Wang, Z.M., and Heymsfield, S.B., 1996, Dual energy X-ray
absorptiometry body composition model: Review of physical concepts. American Journal of
Physiology, 271, 941-951.
BODY COMPOSmON IN PAEDIA TRIes 75
Pietrobelli, A., Wang, Z.M., Formica, c., and Heymsfield, S.B., 1998, Dual energy x-ray
absorptiometry: fat estimation errors due to variation in soft tissue hydration. American Journal of
Physiology, 274, 860-866.
Pietrobelli, A., Wang, Z.M., and Heymsfield, S.B., 1998, Techniques used in measuring human body
composition. Current Opinion in Clinical Nutrition and Metabolic Care, I, 439-448.
Quetelet, L.A.J., 1869. Physique Sociale, Vol 2, C. Muquardt, Brussels.
Rocchini, A.P., 1993, Hemodynamic and cardiac consequences of childhood obesity. In Prevention and
treatment of childhood obesityedited by c.L. Williams and S.Y.S. Kimm.The New York Academy
of Sciences, 3, 46-56.
Roche, A.F., Siervogel, R.M., Chumlea, W.B., and Webb, P., 1981, Grading body fatness from limited
anthropometric data. American Journal of Clinical Nutrition, 34, 2831-38.
Rolland-Cachera, M.-F., 1995, Prediction of adult body composition from infant and child
measurements, In Body composition techniques in health and disease edited by P.W.D. Davies and
T. Cole T. (Cambridge: Cambridge University Press), 100-145.
Rossner, S., 1998, Childhood obesity and adulthood consequences. Acta Paediatrica, 87:1-5.
Sjostrom, L., 1991, A computer tomography based multicomponent body composition techniques and
anthropometric predictions of lean body mass, total and subcutaneous adipose tissue, International
Journal of Obesity, 15, 589-599.
Sjostrom, L., Kvist, H., Cederblad, A., and Tylen, U., 1986, Determination of total adipose tissue and
body fat in women by computed tomography, 40K, and tritium. American Journal of Physiology,
250,736-745.
Slaughter, M.H., Lohman, T.G., Boileau, R.A., Horswill, C.A., Stillman, R.I., Van Loon, M.D., and
Bemben, D.A., 1988, Skinfold equations for estimation of body fatness in children and youth,
Human Biology, 60, 709-723.
Troiano, R.P., and Flegal, K.M., 1998, Overweight children and adolescents: description, epidemiology,
and demographics. Pediatrics, !OI, 497-504.
Ulijaszek, S.1., and Kerr, D.A., 1999, Anthropometric measurement error and the assessment of
nutritional status. British Medical Journal, 82, 165-177.
Wang, Z.M., Heshka, S., Pierson, R.N. Jr., Heymsfield, S.B., 1995, Systematic organization of body
composition methodology: overview with emphasis on component-based methods. American
Journal of Clinical Nutrition, 61, 457-65.
Wang, Z.M., Pierson, R.N. Jr., and Heymsfield, S.B., 1992, The five level model; a new approach to
organizing body composition research. American Journal of Clinical Nutrition, 56, 19-28.
Wagner, D.R., and Heyward, V.H., 1999, Techniques of body composition assessment: a review of
laboratory and field methods. Research Quarterly for Exercise and Sport, 70, 135-149.
Williams, D.P., Going, S.B., Lohman, T.G., Harsha, D.W., Sriinivasan, S.R., Webber, L.S., and
Berenson, G.S., 1992, Body fatness and risk for elevated blood pressure, total cholesterol, and
serum lipoprotein ratios in children and adolescents. American Journal of Public Health, 82, 358-
362.
GENETIC AND ENVIRONMENTAL
FACTORS
CHAPTER 7
1. INTRODUCTION
Obesity is a major health concern throughout the world (e.g., Bouchard, 1994). Body
fat is variously measured through under water weighing or through bioelectric
impedance. While these measures provide accurate direct measurements of total body
fat or percent body fat, often highly correlated surrogates are used in epidemiological
studies, such as the sum of skinfolds measured at multiple sites or the body mass
index (BMI), computed as the weight (kilograms) over the squared height (meters).
Several indices of fat patterning are in wide use: Trunk-to-Extremity Ratio (TER,
computed as the sum of skinfolds measured close to the trunk area over the sum of
skinfolds measured in distal areas) provides a measure of relative fat patterning.
Waist-to-hip ratio and abdominal obesity (visceral plus subcutaneous fat) are also
commonly used. Ironically, the BMI is highly correlated with both total body fat as
well as measures of fat patterning. Accordingly, BMI tends to be used in most
studies because of its simplicity of measurement. Several large family studies are
currently involved in an evaluation of the genes underlying adiposity (e.g., Bouchard
et ai., 1995). Since most family studies include BMI, it provides a good example for
investigating the genetics of fat patterning. Like most quantitative traits and disease-
related risk factors, BMI is a complex trait that involves the action of multiple genes
and environmental effects. In this paper, we will review the general problems
involved in genetic studies of complex traits in general, and review some of the
known results for BMI in particular.
Genetic investigations of simple Mendelian traits have been relatively
straightforward, although investigators are familiar with potential complications
even with simple traits. In contrast, everything about complex traits has been real
difficult, and often even large scale investigations end up with frustrating results
(perhaps all that is simple about complex traits begins and ends with the spelling).
Experienced investigators realise that numerous genes and environmental factors
interact with one another to produce the traits. Not surprisingly, simple-minded
approaches have largely been unsuccessful in finding genes for complex traits and
79
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 79-89.
© 2001 Kluwer Academic Publishers.
80 CHAPTER 7
have generated much controversy with "conflicting" findings from multiple studies.
It is helpful to distinguish between three types of gene effects: major genes, each
with a large effect; oligo genes, each with a moderate effect; and, minor genes or
polygenes, each with a modest effect. For complex traits like fat patterning, there
may not be any major genes at all, and even if there are major genes, their effects are
likely to be mediated by other interacting determinants (other genes and/or
environments). Efforts to date have been relatively more successful in finding major
gene effects, when they exist, but finding genes of small effect sizes (oligo genes and
polygenes) has been very difficult. Numerous large scale studies are underway for
finding genes for hypertension, heart disease, asthma, psychiatric diseases, just to
name a few. Likewise, several groups of investigators are working frantically to map
genes for obesity and other body size phenotypes. Several findings are emerging,
none too compelling. Most often, these studies end up reporting suggestive levels of
evidence.
The effect-sizes of any of the multiple genes involved are likely to be rather
modest. Therefore, methodologies meant for detecting genes with large effects (major
genes) are unlikely to be very successful with complex traits (except in rare
instances), as the experience of recent years has shown. Even though the individual
gene effects may be small, interactions among the genes and environments could
make a substantial contribution to the final manifestation of the trait. Failure to
recognise and accommodate such interactions may often mask the effects of the
individual genes.
A common approach to enhance the power of any study is to utilise larger
sample sizes. The concept of multi-centre genetic studies (e.g., Higgins et aI., 1996)
is rapidly evolving as a means of generating large samples of standardised family
data. Even in pre-planned collaborations of this sort where common protocols are
used and data collection is standardised, one must remain cognisant that the frequency
and distribution of risk factors - both genetic and environmental - may well be
different among the study centres. To pool data from studies that were conducted
independently encompasses even greater challenges as there may be considerable
differences in the sampling strategy, the phenotypic measurement, the particular
genetic markers that were typed, or in the ancillary information available for
classification or phenotypic adjustment. Therefore, sometimes the data may not be
directly poolable, but it may be useful to pool the results from different studies.
Some of these issues have been considered in the development of meta-analytic
methods for pooling results from multiple linkage studies (e.g., Li and Rao, 1996;
Gu et al., 1998b).
For complex traits where we expect etiologic heterogeneity, one wishes to
maximise the signal to noise ratio by analysing the largest possible sample of
families sharing the same predominant etiologic factor(s). Strategies that enable
investigators to sort families into relatively more homogeneous subgroups are
extremely desirable. An approach that holds promise is an application of the
classification and regression trees (CART) methodology, which we refer to as "tree
linkage" method (e.g., Rao, 1998; Shannon et al., 2000; Province et aI., 200la). We
believe that a combination of the lumping (as done by pooling data from multiple
studies or through meta-analysis) and splitting (as done in the tree linkage)
approaches provides an optimal strategy for genetic studies of complex traits.
GENETICS OF COMPLEX TRAITS 81
Gene
~~l~
\( /r
I p I
3. STUDY DESIGN
One can hardly overemphasise that study design is perhaps the single most important
issue in the planning of any genetic study. Some would argue that choice of analysis
methods is less important than a carefully developed study design. Feasibility of the
study, statistical power, and cost-effectiveness all depend critically on the design. It
is important that all the available information about the diseaseltrait (e.g.,
physiology, aetiology etc.) be used fully when decisions are made about the
sampling schemes, sampling units, and analytical methods. More information
should lead to better designs. The major steps involved in a study design are:
definition of the phenotype, sampling unit and the method of sampling, sample size
and power, and cost benefit analysis. Additional issues for genome wide scans
involve various genotyping issues such as the quality of large scale genotyping,
marker density, type of markers, and whether to undertake a linkage scan or an
association scan.
Although definition of the phenotype may at first seem to be a trivial issue,
some thought should be given to whether the current definition of the phenotype,
however expertly done originally, is still the right one to use in gene finding studies.
After all, our goal is one of finding the trait genes, not whether we follow a
traditional or a revolutionary approach. Different definitions of the phenotype do lead
to different results. Certain definitions tend to dwarf the signal while others might at
least have the potential to sharpen, and relatively enhance, the signal. For example,
when studying BMI, it would be prudent to take multiple measurements of both
height and weight. Using the averages will reduce external noise. Likewise, when
studying a disease like essential hypertension, it would be preferable to use early
onset families since they are more likely to be of genetic origin. For phenotypes that
are not highly reproducible, it would seem desirable to study them in smaller family
units rather than in extended pedigrees. Often, end-point phenotypes may be difficult
to define or may not be highly reproducible, but causally-related intermediate
phenotypes may be more easily handled. In such cases, it may be prudent to study
the intermediate phenotypes. Several features of any genetic study are highly
interdependent. The most critical among them are the sampling unit, the sampling
method, and the sample size. One should not be decided independently of the other
two in particular. For genome scans involving complex traits, sib pairs of one type
or another are commonly used in conjunction with model-free methods of analysis.
When using sib pairs, the total number of participants needed in a study can be
minimised by sampling larger sibships as opposed to sampling independent sib pairs
(Todorov et ai., 1997). Other more powerful sampling units such as extremely
GENETICS OF COMPLEX TRAITS 83
discordant (ED) sib pairs (Eaves, 1994; Risch and Zhang, 1995), or extremely
discordant and extremely concordant (EDAC) sib pairs (Gu et at., 1996), can reduce
the sample size even more. Sampling some sibs from above the 90th percentile of a
trait distribution and other sibs from below the 30th percentile appears to provide an
optimum strategy. This includes ED sib pair(s), sib pairs both above the 90th
percentile (high concordant, HC), and sib pairs both below the 30th percentile (low
concordant, LC). The HC is analogous to the affected sib pair (ASP) method.
However, for discrete diseases, an affected individual and an unaffected sib do not
constitute an ED sib pair; they would represent a 'discordant' sib pair, not
necessarily an 'extremely discordant' sib pair. Simply discordant sib pairs do not
constitute a good design. For an extended discussion of study design issues, see Gu
and Rao (2001).
Another crucial issue is the choice of populations from which to sample families
for genetic epidemiological studies of complex phenotypes. It has been argued that
genetically simplified isolates are more informative than diverse admixed populations
for mapping genes for complex phenotypes (Wright et aI., 1999). If strong linkage
disequilibria (LD) exist between marker loci and trait loci in isolated populations,
such an approach should ease the task of gene-mapping. However, it is not clear if
any of the known isolates demonstrate the level of LD needed for gene mapping.
Finally, genetic and environmental heterogeneity may be less in isolated populations
than in conglomerate populations, thus making population isolates more attractive
for gene finding studies using approaches that do not rely heavily on LD. This also
raises a related issue about pooling data from multiple studies. While pooling
increases the sample size and hence the power, it also risks increasing the amount of
heterogeneity. It is therefore highly desirable that splitting the pooled data into
relatively more homogeneous subgroups should be a part of the overall approach to
data pooling, as outlined later.
4. METHODS OF ANALYSIS
Three classes of methods are most commonly used today for linkage analysis: the
model-based classical LOD score method (Morton, 1955); the so-called model-free
relative pair methods (see Elston and Cordell, 2001); and finally the hybrid variance
components methods (Blangero and Almasy, 1997; Province et aI., 2000).
Especially for complex traits, one lacks reasonable trait models and therefore
routine use of the (strongly) model-based LOD score method may not be always
appropriate. This realisation has given rise to the development of alternative methods
that are not based on strong assumptions about the trait inheritance. It is natural to
reason that the existence of a susceptibility gene should lead to an elevated
probability that a pair of affected siblings would inherit the same allele(s) from their
parents. Based on this premise, a class of model-free methods has been developed
based on the sharing of alleles identical by descent (IBD) among relative pairs (see
Elston and Cordell, 2001).
For quantitative traits, the first insightful method was presented by Haseman and
Elston (1972), who took the squared difference of trait values of a sib pair as the
outcome variable, and regressed it on the proportion of alleles shared IBD by the sib
pair using the model E(YjJnj) = BO + BInj- A significantly non-zero negative
84 CHAPTER 7
regression coefficient B1 implies genetic linkage to the marker. Goldgar (1990) and
Amos (1994) used maximum likelihood methods to model directly the covariance
structure of sib pairs and arrived at a variance components method that was more
powerful than the original H-E method. Fulker and Cardon (1994) extended the HE
method to estimate the location of a QTL using flanking markers by applying the
interval mapping method of Goldgar (1990). More recently, Ghosh and Majumder
(2000) have proposed another model-free method for QTL mapping based on
genome-wide scan data on sib pairs. Based on rank correlations of squared differences
of trait values of sib pairs and IBD scores at marker loci and kernel-smoothing
techniques, these authors have proposed a method and have shown that it performs
efficiently under a wide variety of scenarios. As a hybrid of the model-based and
model-free methods, the variance components method combines the strengths of both
methods and provides one of the most powerful methods for linkage analysis of
complex traits. These methods use all the data available within a pedigree, without
excluding subjects with partially missing information or producing redundancy in the
statistics by counting all relative pairs.
Methods for the analysis of associations have undergone many enhancements in
the last decade. Although candidate genes have been the primary focus of association
studies, genome-wide association scans have been proposed as a promising approach
(Risch and Merikangas, 1996). It remains to be seen to what degree the unpredictable
pattern of linkage disequilibrium proves to be a limitation for genome-wide
association scans. While the full promise of this approach remains unclear, partial
association scans in the narrow genomic regions identified by linkage seem to be
particularly attractive. For a comprehensive collection of state-of-the-art methods for
linkage and association analyses, see Rao and Province (2001).
5.1 Meta-analysis
When raw data from multiple studies are not available for pooling, results from
those studies may be pooled using meta-analysis methods. The term "meta-analysis"
is used for a wide variety of statistical procedures developed for summarising results
from multiple studies (see Olkin, 1995 for a review). The application of meta-
analysis techniques to genetic studies began only recently (Li and Rao; 1996; Rice,
1998; Gu et aI., 1998b). For meta-analysis of linkage results, Gu et al. (1998b)
proposed using the proportion of alleles shared IBD at a marker locus by a sib pair
(with specified trait outcomes) as the common effect, and presented methods for
pooling results from model-free sib pair analyses. A random effects model was used
to characterise the among-study variability, and a weighted estimate of the overall
GENETICS OF COMPLEX TRAITS 85
effect and the variance components were given using the weighted least-squares
method. A heterogeneity test was also proposed to assess variability among studies.
For details, see Gu et at. (2001).
Another characteristic feature of complex traits is that the effects of individual loci
often manifest in a battery of correlated traits, and this additional information can be
exploited using appropriate multivariate methods of analysis (e.g., Blangero and
Almasy, 1997; Todorov et aI., 1998; and Province et al., 2000). Full multivariate
methods for simultaneous analysis of multiple traits may be used with greater
power, as argued by Ghosh and Majumder (2001). Other multivariate methods such
as principal component analysis (PCA) and factor analysis are often used to reduce
the dimensionality of data (e.g., Bartholomew, 1987). With respect to genetic
studies, the method of PCA can be used to construct a few "summary phenotypes"
(explaining most of the variance) from a large number of correlated traits, which can
be used in turn for genetic analysis. Alternatively, the method of "principal
components of heritability" (Ott and Rabinowitz 1999) can be directly built into
variance components models such as SEGPATH (Province et at., 2000).
some anthropometric traits such as height is largely genetic in origin, for many
other traits (such as body mass index or BMI, sum of skin folds at multiple sites
etc), the resemblance in families arises from the joint action of genes and
environments (e.g., Rice et at., 1995). As noted earlier, BMI (as estimated by the
body weight in kilograms over the square of height in meters) represents a good
surrogate for both adiposity and regional fat distribution. We shall briefly review
here the current status of the genetics of BML
Familial resemblance of BMI has been investigated extensively by several
investigators in multiple family studies using different analytical methods such as
familial correlations and heritability, segregation analysis, and linkage analysis based
on genome-wide scans with hundreds of anonymous markers. Heritability for BMI
has been estimated around 40% in several studies (e.g., see Rice et at., 1999a; Price
et at., 1994), which has been shown to be remarkably stable over time (Rice et at.,
1999a). That is, about 40% of the variability in BMI can be attributed to the action
of genes and familial environments. Segregation analyses in multiple family studies
have also demonstrated the existence of one or possibly two major genes with
additional sources of familial resemblance (e.g., Price et aZ., 1994; Borecki et at.,
1998; Borecki et at., 1993; Rice et aZ., 1999b). Thus, BMI appears to involve the
joint influences of one or two genes with sufficiently large effects, some minor gene
effects, and possibly some familial environmental effects.
More recently, several genome-wide linkage scans have been undertaken to
localise the genes involved. Chagnon et at. (2001a) have reported linkage evidence
for BMI to several genomic regions in the Quebec Family Study in which
segregation analysis has indicated the existence of one or two major genes (Borecki
et at., 1993). A similar genome scan was also performed in the HERITAGE family
study, and the results are reported in Chagnon et at. (2001b). The pertinent results
for BMI are presented in Table 1.
Table 1. Linkage evidence for BMI from genome-wide scans in two family studies: the
Quebec Family Study (QFS, Chagnon et al. 200Ia) and the HERITAGE Family Study
(HERITAGE, Chagnon et al. 200Ib). Evidence is shown only for those genomic regions
that yielded multipoint LOD scores ;(? 2.0 using the variance components linkage model as
implemented in SEGPATH (Province et al. 2000)
7. DISCUSSION
The study of human disease has fully arrived in the molecular age. The steady and
swift progress of the Human Genome Project has provided molecular tools necessary
for the genetic dissection of complex traits. Yet significant challenges remain. Many
complex traits are also common in the population, accounting for a significant
proportion of the public health burden which emphasises the importance of the
endeavour. We begin the new millennium with challenges, but with it come
enormous opportunities. This is what makes it an exciting time to be a genetic
epidemiologist.
A genetic epidemiologist can hardly be over equipped with tools, and a tool that
works in one case may not work in the next. As experienced investigators realise,
reliance on a single method of analysis, however appropriate it may seem under the
circumstances, is not very optimal. So long as our primary objective is to find genes
for complex traits, we should be willing to consider alternative strategies toward
achieving the objective. In particular, we believe that a "lumping" and "splitting"
strategy can be useful. Lumping involves pooling data and/or results from multiple
studies. Although pooling data may sometimes introduce greater heterogeneity, and
therefore may at first seem counter-intuitive, it is capable of providing substantially
more statistical power when pooling is followed by splitting. Splitting involves
subdividing the aggregate data into multiple homogeneous subgroups, and this may
be done through an application of the CART methodology or through other
multivariate clustering techniques. In this case, separate analyses within subgroups
may actually come with inherently more power on the whole. These strategies would
be particularly useful if, for example, different trait genes or subsets thereof operate
within the subgroups.
88 CHAPTER 7
8. REFERENCES
Amos, C.L, 1994, Robust variance-components approach for assessing genetic linkage in pedigrees.
American Journal of Human Genetics, 54, 535-543.
Bartholomew, D.J., 1987, Latent variable models and factor analysis (Oxford: Oxford University
Press).
B1angero, J., and Almasy, L., 1997, Multipoint oligogenic linkage analysis of quantitative traits. Genetic
Epidemiology, 14,959-964.
Borecki, LB., B1angero, J., Rice, T., Bouchard, C., and Rao, D.C., 1998, Evidence for at least two loci
affecting human fatness. American Journal of Human Genetics, 63, 831-38.
Borecki, 1.8., Bonney, G.E., Rice, T., Bouchard, c., and Rao, D.C., 1993, Influence of genotype-
dependent effects of covariates on the outcome of segregation analysis of the body mass index.
American Journal of Human Genetics, 53, 676-687.
Bouchard, c., 1994, The genetics of obesity (Boca Raton: CRC Press).
Bouchard, C., Leon, A.S., Rao, D.C., Skinner, J.S., Wilmore, J.H., and Gagnon, J., 1995, The
HERITAGE family study: Aims, design, and measurement protocol. Medicine and Science in
Sports and Exercise, 27, 721-729.
Breiman, L., Friedman, J.H., OIshen, R.A., and Stone, C.H., 1984, Classification and Regression Trees
(Belmont, CA: Wadsworth International Group, Inc).
Chagnon, Y.c., Borecki, LB., Perusse, L., Roy, S., Lacaille, M., Ho-Kim, M.A., Chagnon, M., Rice, T.,
Collier, G.R., Rao, D.C. and Bouchard, c., 2001a, Genome-wide search for genes related to
adiposity in the Quebec Family Study. Submitted to Obesity Research.
Chagnon, Y.C., Rice, T., Wilmore, 1.H., Ho-Kim, M.A., Lacaille, M., Pare, C., Bouchard, C., Borecki,
LB., Gagnon, J., Perusse, L., Leon, A.S., Skinner, 1.S., Rao, D.C., and Bouchard, c., 200lb
Genome-wide search for genes related to body composition in sedentary caucasians from the
Heritage Family Study. Pending review in Journal of Applied Physiology.
Eaves, L.l., 1994, Effect of genetic architecture on the power of human linkage studies to resolve the
contribution of quantitative trait loci. Heredity, 72, 175-192.
Elston, R.c., and Cordell, H., 2001, Overview of model-free methods for linkage analysis. In Genetic
Dissection of Complex Traits, edited by D.C. Rao and M.A. Province MA (San Diego: Academic
Press) pp 135-150.
Fulker, D.W., and Cardon, L.R., 1994, A sib-pair approach to interval mapping of Quantitative Trait
Loci. American Journal of Human Genetics, 54, 1092-1103.
Ghosh, S., and Majumder, P.P., 2000, A two-stage variable-stringency semiparametric method for
mapping quantitative-trait loci with the use of genome wide-scan data on sib pairs. American
Journal of Human Genetics, 66,1046-1061.
Ghosh, S., Majumder, and P.P., 2001, Deciphering the genetic architecture of a multivariate phenotype.
Advances in Genetics, 42, 323-347.
Goldgar, D., 1990, Multipoint analysis of human quantitative genetic variation. American Journal of
Human Genetics, 47, 957-967.
Gu, C., and Rao, D.C., 2001, Optimum study designs. In Genetic Dissection of Complex Traits, edited
by D.C. Rao and M.A. Province MA (San Diego: Academic Press) pp. 439-457.
Gu, C., Province, M.A., and Rao, D.C., 2001, Meta-analysis for model-free methods. In Genetic
Dissection of Complex Traits, edited by D.C. Rao and M.A. Province MA (San Diego: Academic
Press), pp 255-272.
GENETICS OF COMPLEX TRAITS 89
Gu, c., Province, M.A., Todorov, A.A., and Rao, D.C., 1998b, Meta-analysis methodology for
combining non-parametric sibpair linkage results: Genetic homogeneity and identical markers.
Genetic Epidemiology, 15,609-626.
Gu, C., Todorov, A.A., and Rao, D.C., 1996, Combining extremely concordant sibpairs with extremely
discordant sibpairs provides a cost-effective way to linkage analysis of QTL. Genetic
Epidemiology, 13,513-533.
Haseman, J.K., and Elston, RC., 1972, The investigation of linkage between a quantitative trait and a
marker locus. Behavior Genetics, 2, 3-19.
Higgins, M., Province, M.A., Heiss, G., Eckfeldt, J., Ellison, RC., Folsom, A.R, Rao, D.C., Sprakfa,
J.M., and Williams, R, 1996, The NHLBI Family Heart Study: Objectives and design. American
Journal of Epidemiology, 143, 1219-1228.
Li, Z., and Rao, D.C., 1996, A random effect model for meta-analysis of mUltiple quantitative sibpair
linkage studies. Genetic Epidemiology, 13,377-383.
Morton, N .E., 1955, Sequential tests for the detection of linkage. American Journal of Human Genetics,
7,277-318.
Olkin, I., 1995, Statistical and theoretical consideration in meta-analysis. Journal of Clinical
Epidemiology, 48,133-146.
Ott, J., and Rabinowitz, D., 1999, A principal-components approach based on heritability for combining
phenotypic information. Human Heredity, 49, \06-11 \.
Price, RA, Charles, M.A., Pettitt, D.J., Knowler, and W.C., 1994, Obesity in Pima Indians: Genetic
segregation analysis of body mass index complicated by temporal increases in obesity. Human
Biology, 66, 251-274.
Province, M.A., Rice, T., Borecki, LB., Gu, C., and Rao, D.C., 2000, A multivariate and multilocus
variance components approach using structural relationships to assess quantitative trait linkage via
SEGPATH. Genetic Epidemiology (In press).
Province, M.A., Shannon, W.A., and Rao DC., 2001a, Classification Methods. In Genetic Dissection of
Complex Traits, edited by D.C. Rao and M.A. Province MA (San Diego: Academic Press), pp 273-
286
Province, M.A., Shannon, W.A., and Rao, D.C., 2001b, Application of tree linkage to hypertensive
sibship data (in preparation).
Rao, D.C., 1998, CAT scans, PET scans, and genomic scans. Genetic Epidemiology, 15, 1-18.
Rao, D.C., and Province, M.A., 2000, Genetic Dissection of Complex Traits (San Diego: Academic
Press).
Rice, J.P., 1998, The role of meta-analysis in linkage studies of complex traits. American Journal of
Medical Genetics, 74, 112-114.
Rice, T., Bouchard, c., Perusse, L., and Rao, D.C., 1995, Familial clustering of multiple measures of
adiposity and fat distribution in the Quebec Family Study: A trivariate analysis of percent body fat,
body mass index, and trunk-to-extremity skinfold ratio. International Journal of Obesity, 19, 902-
908.
Rice, T., Despres, J.P., Daw, E.W., Gagnon, J., Borecki, I.B., Perusse, L., Leon, A.S., Skinner, J.S.,
Wilmore, J.H., Rao, D.C., and Bouchard, c., 1997, Familial resemblance for abdominal visceral fat:
The HERITAGE Family Study. International Journal of Obesity, 21, \024-103 \.
Rice, T., Perusse, L., Bouchard, C., and Rao, D.C., 1999a, Familial aggregation of body mass index and
subcutaneous fat measures in the longitudinal Quebec Family Study. Genetic Epidemiology, 16,
316-334.
Rice, T., Sjostrom, C.D., Perusse, L., Rao, D.C., Sjostrom, L., and Bouchard, C., 1999b, Segregation
analysis of body mass index in a large sample selected for obesity: The Swedish Obese SUbjects
study. Obesity Research, 7, 246-255.
Risch, N., and Merikangas, K., 1996, The future of genetic studies of complex human diseases.
Science, 273, 1516-1517.
Risch, N., and Zhang H, 1995, Extreme discordant sib-pairs for mapping quantitative traits in humans.
Science, 268, 1584-1589.
Shannon, W.A., Province, M.A., and Rao, D.C., 2001, Tree-based recursive partitioning methods for
subdividing sibpairs into relatively more homogeneous subgroups. Genetic Epidemiology, In press.
Todorov, A.A., Province, M.A., Borecki, I.B., and Rao, D.C., 1997, Trade-off between sibship size and
sampling scheme for detecting quantitative trait loci. Human Heredity, 47,1-5.
Todorov, A.A., Vogler, G.P., Gu, c., Province, M.A., Li, Z., Heath, A.C., and Rao, D.C., 1998, Testing
causal hypotheses in multivariate linkage analysis of quantitative traits: General formulation and
application to sibpair data. Genetic Epidemiology, 15,263-278.
Turner, S.T., Boerwinkle, E., and Sing, C.F., 1999, Context-dependent associations of the ACE I/D
polymorphism with blood pressure. Hypertension, 34, 773-778.
Wright, A.F., Carothers, A.D., and Pirastu, M., 1999, Population choice in mapping genes for complex
disease. Nature Genetics, 23, 397-404.
CHAPTER 8
1. INTRODUCTION
The changes of fat distribution during the growth period are the result of changes in
body shape, proportions and body composition and can be modulated by nutrition
and several cultural and socio-economic factors present in human environments. The
distribution of body fat can be studied in different ways, for example, i) from
centripetal fat ratio or trunk extremity ratio, ii) by combining waist-hip ratio and
skinfold ratios (Fiori et at. 2000), iii) by Principal Component Analysis (PCA) on a
set of skinfolds (Mueller and Reid 1979, Ramirez and Mueller 1980), iv) from the
residuals of the regression of each log transformed skinfold on the mean log skinfold
thickness for an individual (Healy and Tanner 1981), v) from ratios of circumferences
and skinfolds (Hattori et al. 1987, Rosique et al. 1994, Rebato et at. 1998) which is
also a heuristic device able to extract fat patterns. However, when input variables are
indices or regressions of the residuals, the size factor is avoided and the shape factor
is first extracted (Baumgartner et at. 1990, Johnston et at. 1991).
One of the earliest attempts to apply a multivariate technique of factor analysis to
study the changes of fat distribution during growth was made by Hammond (1955).
Later on many authors continued to study this basic aspect, often called fat
patterning, by applying PCA, and consequently the literature on this topic has
meanwhile been abundant (Garn and Clark 1976, Johnston 1988, Malina and
Bouchard 1988, Kaplowitz et al. 1988, Deutsch et at. 1985, Mueller 1982, 1985,
1988, Bailey et at. 1985, Baumgartner et at. 1986, Malina et at. 1982, Cameron et
al. 1992, Johnston et al. 1995, Rosique et at. 1994, Robson et al. 1971, Ramirez
and Mueller 1980, Mueller and Wohleb 1981, Demarchi and Marcellino 1999).
Further, to reveal the developmental aspects of fat patterning from the auxological
perspective, normative studies have been found to be increasingly important (Norgan
1987, Johnston 1992, Bouchard et at. 1990). Some of these studies have established
91
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 91-108.
© 2001 Kluwer Academic Publishers.
92 CHAPTER 8
that during the period of adolescence, redistribution of body fat from the extremities
toward the trunk takes place more among the males than among females. The
phenomena has been popularly designated as the Centripetal Fat Distribution (CFD),
and sometimes Centripetal Fat Patterning, which has developed the interests of the
researchers to explain the variety of biological mechanisms involved (Bouchard
1992). Eveleth and Tanner (1990) have stated that subcutaneous fat over the limbs
specially in males, is controlled by different physiological mechanisms from those
regulating subcutaneous fat in thorax particularly during adolescence. Moreover, it
has been recently shown that risk factors for disease association with various types
of fat patterns rise during the period of adolescence (Sarria 1992). Thus generating
more information on the growth in fat patterning during adolescence in particular,
and specifically for the unstudied populations, has been suggested to be necessary for
making proper clinical comparisons (Norgan 1987). Cronk et at. (l983a,b)
emphasised that auxological studies of fat distribution are important in the
peripubertal period when changes in terms of subcutaneous fat in skinfolds are
pronounced.
Genetic factors playa relatively greater role than environment in determining the
fat pattern (Garn 1955, Norgan 1991, Malina et at. 1982, Bouchard 1988, Mueller
and Reid 1979, Selby et at. 1990). However the socio-economic status (SES) is one
of the major environmental determinants. A cross-cultural interpretation of the socio-
economic levels is needed because only in some unwesternised cultures there is a
positive view on fat accumulation in the highest socio-economic level as a sign of
well-being, while in western societies a more slim body is a sign of well-being. For
example, while among the Tokelau migrants and the Dogrib Indians the higher
socio-economic status has been found to be associated with greater CFD (Ramirez
and Mueller 1980, Szathmary and Holt 1983). Further, many workers have reported
that low SES is strongly associated with greater centralisation of subcutaneous fat in
various populations particularly from rural Columbia (Mueller 1986), some ethnic
groups of Guatemala (Bogin and Sullivan 1986, Haffner et at. 1986), Punjabi
adolescents of north India (Johnston et at. 1991), and also other populations,
predominantly from the Western World (Larsson et at. 1989, Gillum 1987, Georges
et at. 1993, Mueller 1986). On the other hand, The National Centre for Health
Statistics Report has shown a significant association of high SES with low CFD
(Georges et at. 1991). Some association between low SES and CFD are also found
to be sex-specific, as shown by Bjorntorp (1988) among the Swedish adult males and
Rebato et at. (1998) on Basque females of Spain. Thus the role of sociocultural
environment on fat patterning has still remained to be one of the important topics of
renewed interest (Georges et at. 1993). It seems that in the western and the most
westernised societies the high SES has a negative view on the accumulation of
central fat and these can influence the frequencies of CFD found among high SES
samples in these kind of populations. The research on this topic can show the
relationship between culture and biology from the Human Ecology perspective.
In spite of scanty information on fat patterning of the children and adults,
recently several recommendations have been made to continue further studies in this
direction of which some may be relevant:
i) As fat pattern has been found to be associated with diseases, ethnic specific
measures of body composition should be developed for the popUlations,
ADDIPOSE TISSUE AND SOCIO-ECONOMIC LEVEL 93
particularly of the multiethnic society, such as in India. This has been given
primary importance for knowing the limits of the anthropometric measurements
and their interpretation in the matters of public health policy instead of the
universal criteria (Solomons and Kumanyika 2000).
ii) Continued efforts should be made towards revealing the effects of various aspects
of socio-economic environment on the relative fat distribution (Wagner and
Heyward 2000).
It may however be noted that due to methodological limitations, it sometimes
becomes difficult to differentiate the effects of social class on fat distribution from
those related to ethnicity (Brown et al. 1992).
From the reviewed literature it becomes quite apparent that, except only a few, all
studies have been conducted on the populations of the Western countries.
Unfortunately studies from the developing countries are found to be virtually lacking
(Cameron et al. 1992). Further, with respect to the population of India in particular,
many studies are now readily available which have particularly investigated the
association between fat distribution and the risk of various diseases (Nirmala Reddy
1998, Das Chaudhuri and Bose 2000). Only two normative studies may be well
quoted of which one has demonstrated the effects of age and SES on fat distribution
(estimated by the PCA) of the north Indian adolescents (Johnston et al. 1991) while
the other one has analysed the fat distribution of the Jat Sikh adolescent boys of
Punjab (Kapoor et al. 1998) aged 11-18 years by using the ratios of circumferences
and skinfolds. The importance of studying the possible effect of age and socio-
economic level on the relative fat patterning, particularly central fat distribution in
the adolescent population of the state of West Bengal could broaden our knowledge
on the relationship between the cultural environment of this important region of
India and the biological characteristics of the Indian population. Thus this paper is
aimed to study the effects of some ecological determinants on central fat distribution
of the Bengali adolescent boys from Calcutta.
collected from the participants. The first set was concerned with the anthropometry
of each subject and the second with the socio-economic and demographic
backgrounds. The subjects were distributed in SES following the classification
provided by the National Classification of Occupations of the fathers (Central
Statistical Organisation 1962).
Subjects wearing only short pant, were measured according to the protocol of the
IBP (Weiner and Lourie 1969). From the set of anthropometric measurements taken
from each subject, four skinfolds (triceps, biceps, subscapular and suprailiac) and calf
circumference were studied in the present paper. However, due to various
organisational constraints, it was not possible to take 5 measurements from all
subjects. Skinfolds were measured by a Lange skinfold calliper (mm) whereas calf
circumference was taken by a steel tape (GPM make). The measurements were taken
by one anthropometrist all through the study on and around the birth dates of the
subjects with 3 days of tolerance between 8 am and 3 pm in the school premises.
The technical errors of the five measurements were calculated from the duplicate
observations (by the formula, TEM = ~ d 2 /2N ) and the values obtained are in the
range of other studies (Ulijaszek and Kerr 1999). The corresponding values are shown
in Table 1.
Measurements n TEM
Triceps skinfold thickness (mm) 18 0.83
Subscapular skinfold thickness (mm) 18 0.29
Biceps skinfold thickness (mm) 11 0.05
Suprailiac skinfold thickness (mm) 11 0.05
Calf circumference (cm) 6 0.08
Expenditure per month in each nuclear family was recorded by interview (accuracy
of the items 50 Rupees) in order to find, as far as possible, an objective correlate of
socio-economic status (SES). Expenditure ranged from 300 to 10,000 Rupees per
month, with a median value of 1,200 Rupees per month and family. As shown in
Table 2, the whole sample was distributed in three SES levels (SESl, SES2 and
SES3) with the help of two cut off points: 1,000 Rupees (the cut off points for the
30th percentile) and 1,500 Rupees (the cut off points for the 70th percentile). The
asymmetry of the chosen cut off points around the mean was preferred instead of
other symmetric cut-off points due to the actual positive skewed distribution of the
expenditure. This procedure could afford comparative subsample sizes of the SES
levels. Moreover, 1,000 to 1,500 Rupees were also meaningful boundaries of this
sample from Calcutta.
In two previous analyses performed on this data set (Dasgupta, unpublished), the
results (2-way ANOV A) revealed statistically significant effect of age and per capita
expenditure level on four of the five traits (three skinfolds and calf circumference)
used in this paper. In addition, the chi-square test revealed significant association
between education and occupation of father with the per capita level of expenditure of
the family. By analysing the dietary data a qualitative difference between the subjects
ADDIPOSE TISSUE AND SOCIO-ECONOMIC LEVEL 95
of the expenditure groups in terms of the consumption of the protein intakes has
been found. The subjects in the higher expenditure group consumed greater frequency
of protein than the subjects belonging in the lower expenditure category.
Principal Component Analysis (PCA) was used to obtain the pattern of fat
distribution. To correct for overall body fat, ratios of relative fat distribution were
included in the analysis (Hattori et al. 1987). The chosen indices have the ability of
maximising the contrast between trunk and extremity fat (Rosique et at. 1994). All
the measures were used in millimetres and transformed with decimal logarithms.
This transformation was able to normalise the distribution. Two consecutive PCA's
were performed in order to test the ability of calf circumference to describe fat
patterning when included in the analysis like skinfolds. Although four skinfolds were
employed in the PCA the performance of the same PCA with the addition of calf
circumference was also analysed.
PCA type-a: the PCA based on only four skinfolds.
PCA type-b: the PCA based on four skinfolds and calf circumference.
The PCA based on four skinfolds was performed using the following indices:
TRI = log triceps/(log subscapular + log suprailiac)
BIC = log biceps/(log subscapular + log suprailiac)
SUB = log subscapular/(log triceps + log biceps)
SUPRA = log suprailiac/(log triceps + log biceps)
The PCA based on four skinfolds and calf circumference was performed using the
following indices:
TRI = log triceps/(log subscapular + log suprailiac)
BIC = log biceps/(log subscapular + log suprailiac)
SUB = log subscapular/(log triceps + log biceps + log calf circumference)
SUPRA = log suprailiac/ (log triceps + log biceps + log calf circumference)
CAL = log calf circumference/(log subscapular + log suprailiac)
After the extraction of three principal components, the factor score for each
individual were obtained to identify patterns of fat distributions. The first factor was
used as an indicator of central body fat distribution. The effect of age on the stability
of the factors was studied by comparing the load of each skinfold index in the
extracted factors from the sample with those obtained in subsequent PCA by age.
The sample was subdivided in two groups, 7 to 12 years and 13 to 16 years. A
separate PCA was used in each age group. The stability of the components across
age was analysed by means of applying subsequent PCA by ages ranges 7 to 12
years and 13 to 16 years.
96 CHAPTER 8
SES differences were investigated based on per month expenditure of the families as
a single socio-economic criteria. Comparison of fat distribution in three different
SES subgroups according to the family expenditure (Table 2) was undertaken by
comparing the fat factor score from both types of PCA-a and PCA-b. A two-way
MANOV A by age and SES has been used in order to test the effects and possible
interactions of age and SES.
Table 2. Distribution of the sample expenditure per month of the nuclear families. The
final sample size and the % of lost sample due to missing values when applying two
different types of peA (a and b) are also shown.
3. RESULTS
Results of the two sets of Principal Component Analysis (PC A type-a and PCA
type-b) performed on the total sample are shown in Table 3. Three extracted
components from type-a (4 skinfolds) and type-b (4 skinfolds and calf circumference)
have explained 99.3% and 93.3% of the variances respectively. In both PCA, the
components showed relationship of shape among variables but not on size. The
skinfolds are associated differently depending on the components and the PCA type (a
or b). However, only correlation values greater than 0.3 in the components have
been considered to be biologically meaningful.
The first component showed a very high correlation with all the indices of both
types of PCA's. The components show a sharp contrast between the limb sites
(triceps, biceps and calf circumference) and the trunk sites (subscapular and
suprailiac). It may however be noted that the corresponding signs of the indices in
two PCA types differ. While in type-a the trunk sites have loaded positively, in
type-b they manifest negative loading. The first component thus have explained
68.8% and 76.6% of the variance and their respective eigenvalues are much above
1.0 (3.1 to 3.4). The component can be designated in various ways, like trunk-
extremity contrast or central-peripheral component. It may further be noted that in
both PCA types suprailiac indices have shown the highest correlational value
ADDIPOSE TISSUE AND SOCIO-ECONOMIC LEVEL 97
followed by calf indices (which is more discernible in the PCA type-b). Among the
extremity sites calf has shown the highest loading (in type-b) followed by triceps
reflecting their greater role than biceps in making the contrast. In the first
component of the PCA type-b, the biceps site has shown relatively lower correlation
of all sites in comparison to the second component.
Table 3. Correlations of the first three principal components extracted with fat
distribution ratios in the total sample (n = 787)
1 2 3 1 2 3
SUPRA 0.89 -0.34 (-0.30)(2) -0.91 (0.20) (-0.01)
BIC -0.88 -0.36 (0.30) 0.59 0.76 (-0.26)
SUB 0.87 0.37 0.33 -0.89 (0.12) (0.17)
TRI -0.87 0.39 (-0.29) 0.82 (0.25) 0.50
CAL 0.89 -0.42 (-0.13)
eigenvalue 3.1 0.5 0.4 3.4 0.9 0.4
% variance 76.6 13.5 9.2 68.8 17.2 7.3
(1) In PCA type a: SUPRA = log suprailiac/(log triceps + log biceps), BIC = log biceps/(log
subscapular + log suprailiac), SUB = log subscapular/(log triceps + log biceps), TRI = log
triceps/(log subscapular + log suprailiac).
In PCA type b: SUPRA = log suprailiac/(log triceps + log biceps + log calf circumference), BIC =
log biceps/(log subscapular + log suprailiac), SUB = log subscapular/(log triceps + log biceps + log
calf circumference), BIC = log biceps/(log subscapular + log suprailiac), SUB = log
subscapular/(log triceps + log biceps + log calf circumference), TRI = log triceps/(log subscapular
+ log suprailiac), CAL = log calf circumference/(log subscapular + log suprailiac).
(2) The coefficients with the lowest contribution to the biological meaning of the components
(coefficients :-; 0.30) are shown in parentheses.
In PCA type-a, biceps and triceps site get almost equal magnitude of loading
with the component 1, but due to the inclusion of calf circumference in the PCA
type-b, in the extremity contrast the contribution of biceps has been minimised.
Thus it becomes apparent that to get the trunk-extremity contrast the role of calf and
triceps should be the primary choice.
The second component of the PCA type-a shows a clear contrast between the
dorsal (subscapular and triceps) and the ventral sites (biceps and suprailiac) which can
also be designated by anterior-posterior component. On the other hand, the second
component of PCA type-b has shown a clear contrast between upper extremity site
(biceps) and lower extremity site (calf circumference). This component has explained
13.5 to 68.5% percent of the variance and their respective eigenvalues are below 1
(0.5 to 0.9). It may be noted that in making the contrast of this component biceps
played a greater role than triceps index (in type-b).
The third component of type-a is correlated with subscapular fat whereas in the
type-b analysis, it is correlated mainly with the triceps fat. It has explained 7.3 to
9.2% of the variance and its eigenvalues are also much lower than 1 (0.4). The third
component has no homogenous meaning across PCA types.
98 CHAPTER 8
With respect to the stability of the components the first component of the PCA
type-a showed a poor stability relative to the PCA of the total sample and with
respect to the patterns of the signs (Table 4). In spite of this, the first component of
the PCA type-a is still a good indicator of the centripetal-peripheral fat pattern and is
consistent across ages. The first component of the PCA type-b showed fair stability
in the pattern of signs but in a lesser degree in the relative strength of the loading.
Table 4. Correlations of the principal components with fat distribution ratios by age
PCA type-b can be viewed as consistent for describing fat patterning during the
growth period due to the stability across age in this Indian male sample.
Mean component scores obtained in both PCA types in relation to SES levels
(family expenditure) are shown in Table 5. In general SES2 seems to have a fat
distribution less centralised than SES levels in the extremes, SES 1 and SES3, as
shown by the mean values of the first components scores of both PCA's. SESI
seems to have more fat backsided distributed (subscapular and triceps) than the other
SES levels as shown by the mean values of the second component scores of PCA
type-a. SES 1 with respect to the other SES levels showed also lower values of the
lower extremity as shown by the mean score obtained in the second component of
the PCA type-b. However the posterior side of the upper extremity had also a
preferential distribution of fat in SES 1 (mean score of the third component in PCA
type-b). Most of these differences were small in magnitude and statistically not
significant.
2 3
mean SD mean SD mean SD
PCA type a 1 172 0.06 0.93 0.15 1.07 0.13 1.05
2 264 -0.10 1.07 0.01 1.00 -0.02 0.99
3 311 0.05 0.97 -0.09 0.95 -0.06 0.98
PCA type b 1 166 -0.05 0.95 -0.08 1.01 0.08 1.01
2 254 0.13 1.03 0.01 1.05 -0.03 1.04
3 302 -0.08 1.00 0.04 0.95 -0.02 0.96
(1) SESI (less than 1,000 Rupees), SES2 (from 1,000 to 1,500 Rupees) and SES3 (more than 1,500
Rupees)
The SES differences were analysed by multivariate design. The vectors of the
components of both PCA types (a and b) were included in a unique MANOVA by
SES (Table 6). The analysis showed significant multivariate differences among the
three subgroups compared (SES1, SES2 and SES3) because the MANOVA yielded a
Wilks' lambda = 0.96 which corresponded to an equivalent F (d.f.: 12, 1428) = 2.25
fairly significant (p < 0.01). However there was no homogeneity of the variances
among the subgroups as shown by the value afforded by Box's M = 134.79 which
corresponded to an equivalent F (d.f.: 42, 984524) = 3.17 (p < 0.001). Only the
second component of the PCA type-a and the first component of the PCA type-b
contributed with significant differences (p < 0.05) to the overall SES differences
found by Wilks' lambda. Moreover an independent multirange comparison test of
Scheffe (= 0.05) was performed (Table 6) with harmonic means due to the size
differences of the subgroups compared, in order to search the subgroups more affected
by SES effects on fat distribution. SES differences in the second component of PCA
type-a were due mainly to the differences between SES 1 and SES3 because the
preferential distribution of fat in the posterior side of the upper body in the lower
SES. SES differences in the first component of PCA type-b were due mainly to
differences between SES2 and the other SES levels, SES 1 and SES3 (Table 6)
because the less central fat distribution of the medium SES level.
Table 7. Results of the two-way MANOVA by age and SES. The component showing
significant value of F contributed to explain the multivariate differences of fat
distribution found by age and SES. However, the significant interaction SES x AGE was
mainly explained by the contribution of the first component of the peA type b, because
of its high F-value, although not significant.
When taking age into account by a two-way MANOVA by SES and age, only
the differences among SES levels in central fat disappeared while the differences in
the antero-posterior pattern of fat distribution remained (Table 7). Age had a greater
contribution than SES to explain the variability in fat distribution as the respective
values of Wilks' lambda showed. The first two components of both PCA types had
ADDIPOSE TISSUE AND SOCIO-ECONOMIC LEVEL 101
o • SES1 • SES1
0.8
-0.2 o SES2 o SES2
• SES3 0.6 • SES3
-0.4
0.8 0.8
(c) Factor 2 (d) Factor 2
0.6 0.6
~
0.4 0.4
0.2 0.2
• SES1
o o o SES2
• SES1 -0.2 • SES3
-0.2
e-- SES2
-0.4 • SES3 -0.4
-0.6 7 -12 13 -16 -0.6 7 -12 13 -16
years years years years
Figure 1. (a) and (c) show the plot of the component scores by age and SES of the first two
fat factors in peA type-a, (b) and (d) show the same plot in peA type-b. The symmetrical
plot of the first factor (I-F) of the peA type-b, was preferred in the graphic because it has
a biological easier interpretation. The other factors do not need transformation.
There were significant interactions age x SES (Wilks' lambda = 0.97) which
corresponded to a significant (p < 0.05) equivalent F (d.f.: 12, 1422) = 1.86. This
two-way interaction was due mainly to the first component of the PCA type-b
because it afforded the highest univariate F-value with respect to the other
components (although not significant F (d.f.: 2, 716) = 2.70, p > 0.05). When
plotting mean component scores by age groups in each SES, the interaction can be
shown also graphically (Figure 1b). Figure 1 shows only the components which
102 CHAPTER 8
4. DISCUSSION
For the last several decades studies on fat distribution (Gam 1955, Norgan 1991) of
the children and adults have gained a considerable importance because of its
significant association with the elevated risk of many diseases (Bjorntorp 1987,
Knowler et al. 1981). The study of the effect of age, sex and SES on fat distribution
in general populations can help to clarify some aspects of the relationship between
fat and risk for some diseases. Differential health risks associated with different types
of body fat distribution has justified the relevancy of studying the stability of the
adipose tissue distribution in general population (Katzmarzyk et al. 1999, Eveleth
and Tanner 1990) and also in populations with high associated risk of disease
(Bouchard et al. 1990, Brenner et al. 1994).
The superiority of the PCA is well known as it utilises all of the information
available in a covariance matrix (Johnston et al. 1991). However, in order to extract
the pattern of fat distribution, researchers have suggested several approaches for
performing the PCA, namely, (i) directly from the skinfold thickness measurements
(Mueller and Reid 1979, Ramirez and Mueller 1980), (ii) from the ratios of the
skinfolds (Cameron et al. 1992, Rebato et al. 1998), (iii) from the residuals of the
regressions of each skinfold on the mean of the log skinfold thickness for an
individual (Healy and Tanner 1981). Hattori et at. (1987) have however found no
major differences in the results by using these approaches. Such a conclusion
becomes more evident when the components extracted in the present study are
compared with those of the Indian study of Johnston et al. (1991) which used the
Healy and Tanner (1981) method.
Although the results of the PCA depends on the variables entered (Baumgartner et
al. 1990, Kaplowitz et al. 1987), the first component of the fat distribution of the
Bengali Boys can unequivocally be identified as the trunk-extremity component as
revealed from the two types of PCA developed in the present research. Since this
component has been obtained from all PCA made so far, it has therefore been coined
as the universal component (Johnston 1992) specific to Homo sapiens (Cameron et
al. 1992). Interchangeably, the component has been called by various names like,
proximal-distal, centripetal, central-peripheral, bipolar, and android-gynoid
continuum, and usually captured most of the variance in comparison to the other
patterns.
The trunk-extremity component of the PCA type-b appears to be more stable and
the contrasts seem to have become more intense due to the inclusion of calf
circumference which presumably has been able to replace the situation when data on
calf skinfold is not available. Thus calf circumference in this study has played a
significant role describing the index of fat of the lower extremity and the contrasts
with the other sides of fat distribution with acceptable performance (68.8%) of the
ADDIPOSE TISSUE AND SOCIO-ECONOMIC LEVEL 103
variance explained by the PCA type-b because the reduction of total variability
explained by the analysis was only 7.8% with respect to PCA type-a. Mueller et al.
(1989) have earlier documented the importance of the circumferential measurements
in extracting the components of fat patterning during the growing years. Further, the
introduction of the leg site has been found to be indispensable for developing the
trunk-extremity contrast (Mueller and Stallones 1981, Mueller and Wohleb 1981). It
is also interesting to note that the trunk-extremity component correlates highest with
the suprailiac index in all PCA's performed. Hence the index may be a sensitive
indicator of this dimorphic fat pattern and should be included as a standard index in
other studies. Coincidentally, among the Cuban infants (Dfaz et al. 1994), the best
association with the first component has been shown by the suprailiac/medial calf
ratio because these two skinfold sites are good indicators to describe the trunk-
extremity contrast.
et al. 1991) and one study on the U.S. children aged 12 years (Deutsch et al. 1985),
the authors found negative loading for the extremity sites and positive for the trunk
sites. On the other hand, Norgan (1987), in his study on the Lufa males obtained the
reverse, i.e., positive for the extremity sites and negative for the trunk sites which
resembles the present results for PCA type-b. Further reversals of the signs of the
first component within a sample is observed to be not unusual (Norgan 1987, Dfaz
et at. 1994). Mueller and Reid (1979) reported reversals in the signs of the first
component between the children and the adolescents. We have also found such
reversals in PCA type-a but without changes in the meaning of the components.
5. REFERENCES
Bailey, S.M., Gershoff, S.N., McGandy, R.B., Nandosita, A., and Tantiwongse, P., 1985, Subcutaneous
fat remodelling in Southeast Asian infants and children. American Journal of Physical
Anthropology,68,123-130.
Baker, P.T., Hunt, E.E., and Sen, T., 1958, The growth and interrelations of skinfo1ds and brachial
tissues in man. American Journal of Physical Anthropology, 16,39-58.
Baumgartner, R.N., Roche, A.F., Guo, S., Cameron Chumlea, W., and Ryan, A.S., 1990, Fat patterning
and centralised obesity in Mexican-American Children in the Hispanic Health and Nutrition
Examination Survey (HHANES 1982-1984). American Journal of Clinical Nutrition, 51, 936S-
943S.
106 CHAPTER 8
Baumgartner, R.N., Roche, A.F., Guo, S., Lohman, T., Boileau, R.A., and Slaughter, M.H., 1986,
Adipose tissue distribution: The stability of Principal Component by sex, ethnicity and maturation
stage. Human Biology, 58, 719-735.
Bjorntorp, P., 1987, Fat patterning and disease: A review. In Human Body Composition and Fat
distribution, edited by N.G. Norgan (Wageningen, Netherlands: Stichting Nederlands Instituut voor
de Voeding). Euro-Nut Report, 8, pp, 201-219.
Bjorntorp, P., 1988, Possible mechanisms of relating fat distribution and metabolism. In Fat distribution
during growth and later health outcomes, edited by C Bouchard and F.E. Johnston (New York:
Alan Riss) pp. 175-19L
Bjorntorp, P., 1996, The regulation of adipose tissue distribution in humans. International Journal of
Obesity, 20, 291-302.
Bogin, B., and Sullivan, T., 1986, Socioeconomic status, sex, age and ethnicity as determinants of body
fat distribution for Guatemalan children. American Journal of Physical Anthropology, 69, 527-535.
Bouchard, C, 1988, Inheritance of human fat distribution. In Fat distribution during growth and later
health outcomes, edited by C Bouchard and F. E. Johnston (New York: Alan R Liss) pp. 103-125.
Bouchard, C, 1992, Heredity and regional fat distribution during growth. In Human Growth: Basic and
clinical aspects, edited by M. Hernandez and 1. Argente. (Amsterdam: Elsevier Science Publishers,
B.V.) pp, 227-243.
Bouchard, C, Bray, G.A., and Hubbard, V.S., 1990, Basic and Clinical aspects of regional fat
distribution. American Journal of Clinical Nutrition, 52, 946-950.
Brenner, J.F., Casey, V.A., Dwyer, J.T., and Bailey, S.M., 1994, The measurement of body fat
distribution using somatotype photographs and computer assisted imaging techniques. Annals of
Human Biology, 21, 23-38.
Brown, D.E., Severance, CJ., Sako, E.K, Chun, D.Y., Young, L.L, and Johnson, J.L., 1992, Growth of
native Hawaiian School children: II. Body Mass Index and skinfold measurements. American
Journal of Human Biology, 4, 433-445.
Cameron, N., Johnston, F.E., Kgamphe, J.S., and Lunz, R, 1992, Body fat patterning in rural South
African Black children. American Journal of Human Biology, 4, 353-364.
Central Statistical Organisation, 1962, Standard Industrial and Occupational Classifications. (New
Delhi: Department of Statistics, Government of India).
Cronk, C.E., Mukherjee, D., and Roche, A.F., 1983a, Changes in triceps and subscapular skinfold
thickness during adolescence. Human Biology, 55, 707-72L
Cronk, CE., Roche, A.F., Kent, R Eichorn, D., and McCammon, R.W., 1983b, Longitudinal trends in
subcutaneous fat thickness during adolescence. American Journal of Physical Anthropology, 61,
197-204.
Das Chaudhuri, A.B., and Bose, K, 2000, A review of studies on central body fat distribution among
Indians (Calcutta: Salmoli Publication).
de Onis, M., Dasgupta, P., Saha, S., Sengupta, D., and Bloessner, M., 2001, NCHS reference and the
growth of the Indian Adolescent boys. American Journal of Clinical Nutrition, (in press).
Demarchi, D.A., and Marcellino, AJ., 1999, Body composition and fat distribution in 10-15 year old
boys from Cordoba, Argentina. Acta Medica Auxologica, 31, 39-44.
Deutsch, M.I, Mueller, W.M., and Malina, RM., 1985, Androgyny in fat patterning associated with
obesity in adolescents and young adults. Annals of Human Biology, 12,275-286.
Diaz, M.E., Montero, M., Wong, I., Moreno, V., Toledo, E.M., and Moreno, R, 1994, Fat Distribution in
Cuban Infants. American Journal of Human Nutrition, 6, 687-692.
Eveleth, P.B., and Tanner, J.M., 1990, Worldwide variation in Human Growth (Cambridge: Cambridge
University Press).
Fiori, G., Facchini, F., Pettener, D., Rimondi, A., Battistini, N., and Bedogni, G., 2000, Relationships
between blood pressure, anthropometric characteristics and blood lipids in high and low altitude
popUlations from central Asia. Annals of Human Biology, 27, 19-28.
Fry, E.I, Chang, KS.F, Lee, M.M.C, and Ng, CK, 1965, The amount and distribution of subcutaneous
tissue in Southern Chinese children from Hong Kong. American Journal of Physical Anthropology,
23,69-80.
Gam, S.M., 1955, Relative fat patterning: an individual characteristic. Human Biology, 27, 75-89.
Gam, S.M., and Clark, D.C, 1976, Trends in fatness and the origins of obesity. Metabolism, 57, 443-
456.
Gam, S.M., and Haskell, J.A., 1959, Fat changes during adolescence. Science, 129, 1615-1616.
Georges, E., Mueller, W.H., and Wear, M.L., 1991, Body fat distribution: Association with
socioeconomic status in the Hispanic Health and Nutrition Examination Survey. American Journal
of Human Biology, 3, 489-50L
Georges, E., Mueller, W.H., and Wear, M.L., 1993, Body fat distribution in men and women of the
Hispanic health and nutrition examination survey of the United States: Association with behavioural
variables. Annals of Human Biology, 20, 275-29L
Gillum, F., 1987, The association of the ratio of waist to hip girth with blood pressure, serum cholesterol
and uric acid in children and youths aged 6-17 years. Journal of Chronic disease, 40, 413-420.
ADDIPOSE TISSUE AND SOCIO-ECONOMIC LEVEL 107
Haffner, S.M., Stern, M.P., Hazuda, H.P., Rosenthal, M., Knapp, J.A., and Malina, RM, 1986, Role of
obesity and fat distribution in non-insulin dependant diabetes mellitus in Mexican-American and
non- Hispanic Whites. Diabetes Care, 9, 153-16l.
Hammond, W.A., 1955, Measurement and interpretation of subcutaneous fat, with norms for children
and young adult males. British Journal of Preventive and Social Medicine, 9, 201-211.
Hattori, K., Becque, M.D., Katch, V.L., Rocchini, A.P., Boileau, RA, Siaughtere, M.H., and Lohman,
T.G., 1987, Fat patterning of adolescents. Annals of Human Biology, 14,23-28.
Healy, M.J.R, and Tanner, J.M., 1981, Size and Shape in relation to growth and form. Symposia of the
Zoological Society of London, 46, 19-32.
Indech, G.D., Sanjeev, I., and Johnston, F.E., 1991, Age, sex and socioeconomic correlates of fat
patterning among adults from Chandigarh zone of Northwest India. American Journal of Human
Biology, 18,463-470.
Johnston, F.E., 1988, Sex differences in fat patterning in children and Youth. In Fat Distribution During
Growth and Later Health Outcomes, edited by C. Bouchard and F.E. Johnston (New York: Alan R
Liss) pp, 85-102.
Johnston, F.E., 1992, Developmental aspects of fat patterning. In Human Growth: Basic and clinical
aspects, edited by M. Hernandez and J. Argente (Amsterdam: Elsevier Science Publishers, B.V.)
pp,217-226.
Johnston, F.E., Heath, B.H., and Shoup, RF., 1995, Patrones de distribucion de la grasa en cuatro
muestras de jovenes de 12-18 anos de edad. Revista Espanola de Antropologfa Biol6gica, 16, 69-
83.
Johnston, F.E., Sanjeev, 1., and Indech, G.D., 1991, Fatness and fat patterning in 12-17 year old youth
from the Chandigarh zone of Northwest India. American Journal of Human Biology, 3, 587-597.
Kaplowitz, H.J, Mueller, WH., Selwyn, B.1., Malina, RM., Bailey, D.A, and Mirwald, R.L., 1987,
Sensitivities, specificities and positive predictive values of simple indices of body fat distribution.
Human Biology, 39, 809-825.
Kaplowitz, H.1., Wild, K.A., Mueller, W.H., Decker, M., and Tanner, J.M., 1988, Serial and parent-
child changes in components of body fat distribution and fatness in children from the London
Longitudinal Growth study, ages, two to eighteen years. Human Biology, 60, 739-758.
Kapoor, S., Patra, P.K. Sandhu, S., and Kapoor, A.K., 1998, Fatness and its distribution pattern among
Jat Sikhs. Journal of the Indian Anthropological Society, 33, 223-228.
Katzmarzyk P., Perusse, L., Malina, R., and Bouchard, e., 1999, Seven year stability of indicators of
obesity and adipose tissue distribution in the Canadian population. American Journal of Clinical
Nutrition, 69, 1123-1129.
Knowler, W.C., Pettitt, D.1., Savage, PJ., and Bennet, P.H., 1981, Diabetes in Pima Indian: contributions
of obesity and parental diabetes. American Journal of Epidemiology, 113, 144-156.
Larsson, B., Seidell, J., Svarsuud, K., Welin, L., Tibblin, G., and Bjorntorp, P., 1989, Obesity, adipose
tissue distribution and health in men. The study of men born in 1913. Appetite, 13,37-44.
Malina, R.M., 1996, Regional Body Composition: age, sex and ethnic variation. In Human Body
Composition, edited by A. Roche, S.B. Heymsfield and T.G. Lohman. (Champaign, IL: Human
Kinetics) pp, 217-255.
Malina, R.M., and Bouchard, e., 1988, Subcutaneous fat distribution during growth. In Fat Distribution
During Growth and Later Health Outcomes, edited bye. Bouchard and F.E. Johnston (New York:
Alan R. Liss) pp, 63-84.
Malina, R.M., Koziel, S., and Bielicki, T., 1999, Variation in Subcutaneous Adipose tissue distribution
associated with age, sex and maturation. American Journal of Human Biology, II, 189-200.
Malina, RM., Mueller, W.H., Bouchard, e., Shoup, R.F., and Lariviere, G., 1982, Fatness and fat
patterning among athletes at the Montreal Olympic Games, 1976. Medicine Science and Sports
Exercise, 14, 445-452.
Mueller, W.H, Marbella, A., Harrist, RB, Kaplowitz, H.A., Grunbaum, J.A., and Lebarthe, D.R., 1989,
Body circumferences as alternatives to skinfold measures of body fat distribution in children.
Annals of Human Biology, 16,49-506.
Mueller, W.H., 1982, The changes with age of the anatomical distribution of fat. Social Science in
Medicine, 16, 191-196.
Mueller, W.H., 1985, The biology of human fat patterning. In Human Body Composition and Fat
Distribution, edited by N. Norgan (London: Euro-Nut) pp, 159-174.
Mueller, W.H., 1986, Environmental sensitivity of different skinfold sites. Human Biology, 58, 499-506.
Mueller, W.H., 1988, Ethnic differences in fat distribution during growth. In Fat Distribution During
Growth and Later Health Outcomes, edited by e. Bouchard and F.E. Johnston (New York: Alan R
Liss) pp, 127-145.
Mueller, W.H., and Reid, R.M., 1979, A multivariate analysis of fatness and relative fat patterning.
American Journal of Physical Anthropology, 50, 199-208.
Mueller, W.H., and Wohleb, J.e., 1981, Anatomical distribution of subcutaneous fat and its description
by multivariate methods: How valid are principal components? American journal of Physical
anthropology, 54, 25 -35.
108 CHAPTER 8
Mueller, W.H., Deutsch. M.l., Malina, RM., Bailey, D.A., and Mirwald, R.L., 1986, Subcutaneous fat
topography: Age changes and Relationship to Cardiovascular fitness in Canadians. Human Biology,
58, 955-973.
Mueller, W.M., and Stallones, L., 1981, Anatomical distribution of subcutaneous fat: skinfold sites and
construction of indices. Human Biology, 53, 321-335.
Nirmala Reddy, B., 1998, Blood Pressure and adiposity: a comparative study of socioeconomically
diverse group of Andhra Pradesh, India. American Journal of Human Biology, 10, 5-21.
Norgan, N., 1987, Fat patterning in Papua New Guineans. Effects of age, sex and acculturation.
American Journal of Physical Anthropology, 74,385-392.
Norgan, N., 1991, Anthropometric assessment of body fat and fatness. In Anthropometric Assessment
of Nutritional Status, edited by Himes, W.H. (New York: WiJey-Liss Inc.) pp, 197-212.
Ramirez, M.E., and Mueller, W.H., 1980, The development of obesity fat patterning in Tokelau
children. Human Biology, 52, 675-687.
Rebato, E., Salces, 1., San Martin, L., and Rosique, J., 1998, Fat distribution in relation to sex and
socioeconomic status in children 4-19 years. American Journal of Human Biology, 19, 799-806.
Risley, H.H., 1891, Tribes and Castes of Bengal. Ethnographic Glossary I and 2. (Calcutta: Bengal
Secretariat Press).
Risley, H.H., 1908, The people of India. (London: Thacker, Sprinker and Co.).
Robson, lRK., Bazin, M., and Soderstrom, R, 1971, Ethnic differences in skinfold thickness. American
Journal of Clinical nutrition, 24, 864-868.
Roemmich, IN., and Rogol, A.D., 1999, Hormonal Changes during Puberty and their relationship to fat
distribution. American Journal of Human Biology, 11,209-224.
Rolland-Cachera, M.F., Bellisle, F., Deheeger, M., Pergquignot, F., and Sempe, M., 1990, Influence of
body fat distribution during childhood: and body fat distribution in adulthood: A two decade follow-
up study. International Journal of. Obesity, 14,473-481.
Rosique, l, Rebato, E., Gonzalez, A., and Pacheco, J.L., 1994, Somatotype related to centripetal fat
patterning of 8-to 19 -year-old Basque boys and girls. American Journal of Human Biology, 6, 171-
181.
Sarria, A., 1992, Methods for assessing fat patterning in children In Human Growth: Basic and Clinical
Aspects, edited by M. Hernandez and J Argente (Amsterdam: Elsevier Science Publishers, B.V.)
pp, 233-244.
Selby, J.V., Newman, B., Quesenberry, Ch.P., Fabsitz, RR, Carmelli, D., Meaney, F.J., and Slemenda,
Ch., 1990, Genetic and Behavioural influences on body fat distribution. International Journal of
Obesity, 14,593-602.
Solomons, N.W., and Kumanyika, S., 2000, Implications of racial distinctions for body composition and
its diagnostic assessment. American Journal of Clinical Nutrition, 71,1387-1389.
Stolz, H.R, and Stolz, L.M., 1951, Somatic development of Adolescent Boys. A study of the growth of
boys during the second decade of life. New York, Macmillan.
Susanne, c., Rebato, E., Hauspie, R.C., Vercauteren, M., Sa ices, I., San Martin, L., and Rosique, I.,
2000, A review of the relationship between nutrition and some growth and Development data. In
Puberty: Variability of changes and Complexity of factors, edited by Bodzsar, E.B., Susanne, C. and
Prokopec, M. (Budapest: Eotovos University Press).
Szathmary, E.J.E., and Holt, N., 1983, Hyperglycemia in Dogrib Indians of the Northwest Territories,
Canada: Association with age and a centripetal distribution of body fat. Human Biology, 55, 493-
515.
Ulijaszek, S., and Kerr, D.A., 1999, Anthroppometric Measurement error and the assessment of
nutritional status. British Journal of Nutrition, 82,165-177.
Wagner, D.R, and Heyward, V.H., 2000, Measures of body composition in blacks and whites: A
comparative review. American Journal of Clinical Nutrition, 71, 1392-1402.
Weiner, J.S., and Lourie, I.A., 1969, Human Biology: A guide to field methods. International Biological
Programme. Handbook, 9 (Oxford: Blackwell Scientific Publications).
CHAPTER 9
T. BROWN, G. TOWNSEND
Dental School, The University of Adelaide, Adelaide, South Australia
children enrolled in the study. Special attention was given to the changing
environment as the Aboriginal community gradually adopted a lifestyle markedly
different from their earlier hunter-gatherer existence (Campbell and Barrett 1953,
Barrett 1958).
-
NORTHERN
-<
.....J TERRITORY
-<~
E- O
rJ)
Z
;:J
-< -<
.....J
~1
rJ)
Yuendumu Reserve Z
tJ.J
tJ.J
tJ.J
r . Al ice Spdng'
~
0
~ ~
SOUTH AUSTRALIA
In 1961 the scope of the study was extended to include a wider range of
observations on the Yuendumu subjects including measurements of the face and
skull and selected general body features. Computer-based methods were developed to
DENTOFACIAL MORPHOLOGY, GROWTH AND GENETICS III
deal with the accumulation of study material which included serial somatometric
data, observations from cephalometric and hand-wrist radiographs, measurements
from dental casts and, of particular importance, genealogical records of Yuendumu
families (Barrett, Brown and Fanning 1965, Barrett, Brown and MacDonald 1965,
Barrett 1969). The overall objective of the researchers was to use the Yuendumu data
in correlated studies of morphological features, functional relations and patterns of
growth and development. Phase one was concerned primarily with descriptive and
comparative aspects of the physical and dental characteristics in order to provide
base-line references for later use. Phase two consisted of correlative analyses where
the dentition and craniofacial structures were considered in conjunction with general
body growth and skeletal maturation. Phase three, the most recent, was initiated by
Townsend (1976) to utilise the genealogical records. The relative geographic
isolation of the Yuendumu community and the practice of polygyny provided the
unique opportunity for genetic studies of variation in the metric and non-metric
dental characters (Townsend and Brown 1978, Townsend 1980, Townsend and
Brown 1981).
Annual trips to Yuendumu continued until 1971 resulting in the extensive
collection of records and observations, housed in the Dental School at The
University of Adelaide, that still provide material for study. To date over 300
publications have arisen from the project, including many by visiting researchers
from Sweden, Denmark, Finland, England, United States of America, Israel, Japan
and India. As the investigation continues, its uniqueness and many advantages are
being reinforced, particularly for the study of the nature and causes of variations in
dental maturation and general physical growth and development.
and Townsend 1980a). This surplus space assists to minimise crowded dental
arches as the permanent canine and premolars emerge into occlusal alignment.
Figure 2. Models of the maxillary dental arches of an Australian Aboriginal (left) and
a white Australian male (right), highlighting marked differences in both size and arch
shape.
other groups in the average number of teeth emerged. The relative delay in primary
tooth emergence may be associated with a prolonged period of breast feeding or it
may be a reflection of the severe illnesses suffered by many of the infants at
Yuendumu.
Emergence of the permanent teeth is conveniently considered in two phases:
phase one includes emergence of the first 12 teeth, that is all incisors and first
molars, while phase two includes emergence of the canines, premolars, second and
third molars. The average sequence of emergence in the Aboriginal children did not
differ greatly from that recorded for other groups but the sequence of the mandibular
second premolar and second molar was variable, either tooth taking priority,
suggesting that the emergence sequence of this tooth pair may be polymorphic
(Barrett, Brown and Cellier 1964).
Phase one was complete when 12 permanent teeth were present, a stage that
extended on. average from 8.5 to 9.9 years in boys and from 8.1 to 9.1 years in girls.
After this plateau stage, phase two commenced with emergence of the first premolar
or canine, whichever was the earlier. All teeth except third molars were present at
about 11.5 and 11.1 years in Aboriginal boys and girls respectively. The quiescent
plateau period, with no teeth emerging, tended to be of shorter duration in the
Aboriginal children compared with several other groups, l.47 years in boys and
0.99 years in girls (Brown 1978). Compared with Australian children of European
descent, the Aboriginal children tended to be advanced during phase two. Thus it
appears that tooth emergence in the Yuendumu children is a process that commences
relatively late but finishes earlier than in some European populations. The entire
process extended from about 1 year when the primary incisors emerge until about 16
years with emergence of the third molars. The low prevalence of minor tooth
crowding and other more severe malocclusions in this group is evidence that the
emergence process is relatively unimpeded.
The association between the paths of tooth emergence and the pattern of facial
growth has been documented by Bjork and Skieller (1972). In this respect
compensatory dento-alveolar growth is an important biological mechanism acting
during development of the dental arches and establishment of dental occlusion. The
pattern of craniofacial growth was studied in an Aboriginal boy between 7 and 18
years of age using cephalometric radiographs (Bjork, Brown and Skieller 1984). The
structural method for superimposing the serial radiographs as outlined by Bjork and
Skieller (1972) was used for the analysis. In this boy the entire dental arch migrated
anteriorly and occlusally during the growth period. Anterior migration was 8 mm in
the region of the first permanent molar in both mandibular and maxillary arches. The
anterior teeth migrated less, 6 mm in the mandible and 5 mm in the maxilla. This
anterior migration of the dental arch, which was considerably more than that
recorded for Danish subjects, was accompanied by a shortening of the dental arches
and increasing alveolar prognathism. As a consequence there was ample space for
emergence of the third molars, a condition not always present in the Caucasian
dentition.
The natural function of the dentition and dental occlusion of Australian
Aborigines was described by Barrett (1958) who emphasised the dynamic and
continuously changing nature of occlusal relationships in this group. According to
Barrett's concepts, the dentition underwent a wear-in and a wear-out phase. During
the wear-in phase, continuing wear on the occlusal tooth surfaces, conditioned by
DENTOFACIAL MORPHOLOGY, GROWTH AND GENETICS 115
For example, in the Aboriginal boys, peak growth velocity in maxillary length,
mandibular length, upper face height and total face height occurred at 13.5, l3.8,
l3.0 and l3.8 years respectively. Relationships in the girls followed a similar
pattern. Correlations between the timing of peak adolescent growth in stature,
mandibular length and facial height were high, ranging from 0.78 to 0.84.
Of the skeletal events studied, it was found that initial ossification of the
pisiform, hook of the hamate and the ulna metacarpophalangeal sesamoid of the first
finger were clustered around the time of peak growth velocity in stature and facial
dimensions as shown in figure 4. Ossification of the pisiform and initial ossification
of the hamate occurred, on average, between 1.2 and 1.7 years before peak stature
velocity whereas late ossification of the hamate hook and the sesamoid occurred at
about the same time as peak growth velocity. For diagnosis and treatment planning
in orthodontics, the timing of these ossification events form a useful guide to a
child's general and craniofacial growth status, particularly if there is reason to
suspect that chronological and skeletal ages differ significantly (Grave and Brown
1979).
10
..
to
~8
e
(.;
...
1
'uQ
- 6
-Z,.
~4
'OJ 1. Pisiform
::t
2. Hamate-I
3. Hamate- 2
4. Sesamoid
"
9 10 12 13 14 15 Ifl
Age - years
various dento-facial features between half-siblings, who have the same father but
different mothers, as well as full-siblings. This unusual population structure enables
insights to be gained into the influence of genetic factors on developing structures,
but also the role of maternal factors that may be operating pre- and post-natally. We
have concentrated mainly on analysing genetic influences on tooth-size variability in
the Yuendumu Aborigines, but we have also attempted to clarify the importance of
genetic effects on morphological crown features, such as Carabelli trait.
Heritability estimates indicate the proportion of observed variation that can be
attributed to genetic influences. Caution is needed in interpreting these values
because they only relate to the population under examination at a given time, and
they may vary for the same population over time, or between different populations
for the same feature. Our approach to estimating heritabilities for tooth size in the
Yuendumu population was based on analysis of variance, with the assumption that
tooth size is controlled by a multifactorial form of inheritance, with both genetic and
environmental components (Townsend and Brown 1978, Townsend, 1980).
Table I shows the results of partitioning tooth-size variation into additive
genetic, common environmental and unique environmental components. Narrow-
sense heritability estimates (V A) of around 60% were found for both deciduous and
permanent dentitions, but common environmental influences (VEe) or maternal
effects appeared to be more important in determining variation of deciduous teeth
(around 15%) than permanent (around 6%). This finding is consistent with the
known timing of development of the two dentitions: deciduous tooth crowns form
mainly pre-natally, whereas permanent teeth only commence to calcify at birth.
Deciduous
mesiodistal 50 19 32
buccolingual 66 12 23
both 58 15 27
Permanent
mesiodistal 63 4 33
buccolingual 66 8 26
both 64 6 30
It is interesting that our more recent studies of tooth size variability in white
Australian twins have yielded higher heritability estimates, most being over 80%
(Dempsey, Townsend and Martin 1999).
As mentioned earlier, care is needed in comparing heritability estimates derived
from different populations, but it is possible that the relatively low estimates for
Aborigines reflect a period of time when the environment was exerting considerable
DENTOFACIAL MORPHOLOGY, GROWTH AND GENETICS 119
The availability of serial dental models for many of the Aboriginal children has
enabled us to examine the expression of various morphological crown features, such
as Carabelli trait, in both the deciduous and permanent dentitions of the same
individuals, as shown in figure 5. Assuming that the genetic control of Carabelli
trait is similar in both dentitions, any differences in expression presumably result
from the influence of environmental factors during the period of crown formation.
Similarly, any differences in expression of dental traits between right and left sides of
individuals must reflect developmental disturbances, if it is assumed that genetic
effects on both sides are the same. We have noted that Carabelli trait tends to occur
120 CHAPTER 9
bilaterally with symmetrical expression in both dentitions. Although the trait was
found relatively often in the deciduous dentition but not the permanent, the reverse
situation was very rare (Townsend and Brown 1981). This suggests that the
genotype may be reflected more faithfully in the deciduous dentition. Reduced
penetrance in the permanent dentition could result from environmental influences
operating during its longer period of development. The deciduous dentition is often
described as being more conservative in evolutionary terms than the permanent, and
it would seem that more genetic studies of the former are warranted, particularly as
there is a move in clinical dentistry to deal with developmental disturbances in the
dentitions of growing children at younger ages.
Acknowledgements. The growth study was supported by grants from the National
Institute of Health, Bethesda, Maryland, the National Health and Medical Research
Council, Canberra, and the University of Adelaide. We are indebted to the
Aboriginal children and adults who participated in the study and to the late Rev.
and Mrs Tom Fleming, Baptist missionaries of Yuendumu, for their continuing
assistance during 20 years of field visits.
7. REFERENCES
Abbie, A.A., 1957, Metrical characters of a Central Australian tribe. Oceania, 27, 220-243.
Barrett, M.1., 1958, Dental observations on Australian Aborigines: continuously changing functional
occlusion. Australian Dental Journal, 3, 39-52.
Barrett, M.1., 1969, Functioning occlusion. Annals of the Australian College of Dental Surgeons, 2, 68-
80.
Barrett, M.1., 1976, Dental Observations on Australian Aborigines: Collected Papers and Reports 1953-
1973. (Faculty of Dentistry: The University of Adelaide, Adelaide).
Barrett, M.1., and Brown, T., 1966, Eruption of deciduous teeth in Australian Aborigines. Australian
Dental Journal, 11,43-50.
Barrett, M.1., Brown, T., and Cellier, K.M., 1964, Tooth eruption sequence in a tribe of Central
Australian Aborigines. American Journal of Physical Anthropology, 22, 79-89.
Barrett, M.J., Brown, T., and Fanning, E.A., 1965, A long-term study of the dental and craniofacial
characteristics of a tribe of Central Australian Aborigines. Australian Dental Journal, 10, 63-68.
Barrett, M.1., Brown, T., and Luke, J.I., 1963, Dental observations on Australian Aborigines -
mesiodistal crown diameters of deciduous teeth. Australian Dental Journal, 8, 299-302.
Barrett, M.J., Brown, T., and MacDonald, M.R., 1963a, Dental observations on Australian Aborigines:
mesiodistal crown diameters of permanent teeth. Australian Dental Journal, 8, 150-155.
Barrett, M.1., Brown, T., and MacDonald, M.R., 1963b, Dental observations on Australian Aborigines:
roentgenographic study of prognathism. Australian Dental Journal, 8, 418-427.
Barrett, M.1., Brown, T., and MacDonald, M.R., 1965, Size of dental arches in a tribe of Central
Australian Aborigines. Journal of Dental Research, 44, 912-920.
Barrett, M.1., Brown, T., Arato, G., and Ozols, LV., 1964, Dental observations on Australian
Aborigines: buccolingual crown diameters of deciduous and permanent teeth. Australian Dental
Journal, 9, 280-285.
Bj6rk, A., and Skieller, V., 1972, Facial development and tooth eruption. An implant study at the age of
puberty. American Journal of Orthodontics, 62, 339-383.
Bj6rk, A., Brown, T., and Skieller, V., 1984, Comparison of craniofacial growth in an Australian
Aboriginal and Danes, illustrated by longitudinal cephalometric analysis. European Orthodontic
Journal, 6,1-14.
Brace, C.L., 1980, Australian tooth-size clines and the death of a stereotype. Current Anthropology, 21,
141-164.
Brown, P., 1992, Post-Pleistocene change in Australian Aboriginal tooth size: Dental reduction or
relative expansion? In T Brown and S Molnar (eds.) Craniofacial Variation in Pacific Populations
(Adelaide: Anthropology and Genetics Laboratory, The University of Adelaide), p.33-51.
Brown, T., 1976, Head size increases in Australian Aborigines. An example of skeletal plasticity. In R
L Kirk and A G Thome (eds.) The Origin of the Australians (Canberra: Australian Institute of
Aboriginal Studies), pp. 195-209.
Brown, T., 1978, Tooth emergence in Australian Aboriginals. Annals of Human Biology, 5, 41-54.
Brown, T., and Barrett, M.1., 1971, Growth in Central Australian Aborigines: stature. Medical Journal
of Australia, 2, 29-33.
Brown, T., and Barrett, M.1., 1972, Growth in Central Australian Aborigines: weight. Medical Journal
of Australia, 2, 999-1002.
Brown, T., and Grave, K.C., 1976, Skeletal maturation in Australian Aborigines. Australian Paediatric
Journal, 12,24-30.
Brown, T., and Townsend, G.c., 1982, Adolescent growth in height of Australian Aboriginals analysed
by the Preece-Baines function: a longitudinal study. Annals of Human Biology, 9, 495-505
122 CHAPTER 9
Brown, T., Abbott, A., and Burgess, V.B., 1987, Longitudinal study of dental arch relationships in
Australian Aboriginals with reference to alternate intercuspation. American Journal of Physical
Anthropology, 72, 49-57.
Brown, T., Barrett, MJ., and Grave, K.C., 1971, Facial growth and skeletal maturation at adolescence.
Tandlregebladet, 75, 1221-1222.
Brown, T., Margetts, B., and Townsend, G.C., 1980a, Comparison of mesiodistal crown diameters of the
deciduous and permanent teeth in Australian Aboriginals. Australian Dental Journal, 25, 28-33
Brown, T., Margetts, B., and Townsend, G.c., 1980b, Correlations between crown diameters of the
deciduous and permanent teeth of Australian Aboriginals. Australian Dental Journal, 25, 219-223.
Campbell, T.D., 1939, Dental conditions of Aborigines compared with civilized groups. Mankind, 2,
228-229.
Campbell, T.D., and Barrett, MJ., 1953, Dental observations on Australian Aborigines: a changing
environment and food pattern. Australian Journal of Dentistry, 57,1-6.
Campbell, T.D., Gray, J.H., and Hackett, CJ., 1936, Physical anthropology of the Aborigines of central
Australia. Part 1. Anthropometry. Oceania, 7, 106-139.
Corruccini, RS., 1999, How Anthropology Informs the Orthodontic Diagnosis of Malocclusion's Causes
(Lewiston: The Edwin Mellen Press).
Corruccini, RS., Townsend, G.C., and Brown, T., 1990, Occlusal variation in Australian Aboriginals.
American Journal of Physical Anthropology, 82, 257-265.
Dempsey, PJ., Townsend, G.c., and Martin, N.G., 1999, Insights into the genetic basis of human dental
variation from statistical modelling analyses. Perspectives in Human Biology, 4(3), 9-17.
Grave, K.C., and Brown, T., 1976, Skeletal ossification and the adolescent growth spurt. American
Journal of Orthodontics, 69, 611-619.
Grave, K.C., and Brown, T., 1979, Carpal radiographs in orthodontic treatment. American Journal of
Orthodontics, 75, 27-45.
Greulich, W.w., and Pyle, S.l., 1959, Radiographic Atlas of Skeletal Development of the Hand and
Wrist. 2nd ed. (Stanford: Stanford University Press).
Lewin, T., and Hedegard , B., 1971, Secular changes in craniofacial dimensions of adult Skolt Lapps:
studies on population and family levels within a genetic isolate group. Suomen
Hammaslaakariseuran Toimituksia, 67,171-183.
Middleton, MR, and Francis, S.H., 1976, Yuendumu and its Children. Life and Health on an Aboriginal
Settlement. (Canberra: Australian Government Publishing Service).
Miller, P.S., 1970, Secular changes among the Western Apache. American Journal of Physical
Anthropology, 33, 197-206.
Moodie, P.M., 1973, Aboriginal Health. (Canberra: Australian National University Press), pp. 181-187.
Preece, M.A., and Baines, MJ., 1978, A new family of mathematical models describing the human
growth curve. Annals of Human Biology, 5, 1-24.
Richards, L.C., and Brown, T., 1986, Development of the helicoidal plane. Human Evolution, 1, 385-
398.
Solow, B., 1969, Automatic processing of growth data. Angle Orthodontist, 39,186-197.
Townsend, G.C., 1976, Tooth Size Variability in Australian Aboriginals: A Descriptive and Genetic
Study. Ph.D Thesis, The University of Adelaide.
Townsend, G.c., 1980, Heritability of deciduous tooth size in Australian Aboriginals. American Journal
of Physical Anthropology, 53, 297-300.
Townsend, G.c., and Brown, T., 1978, Inheritance of tooth size in Australian Aboriginals. American
Journal of Physical Anthropology, 48, 305-314.
Townsend, G.C., and Brown, T., 1978, Heritability of permanent tooth size. American Journal of
Physical Anthropology, 49, 497-504.
Townsend, G.C., and Brown, T., 1979, Tooth size characteristics of Australian Aborigines. Occasional
Papers in Human Biology, 1, 17-38 (Canberra: Australian Institute of Aboriginal Studies).
Townsend, G.C., and Brown, T., 1981, The Carabelli trait in Australian Aboriginal dentition. Archives
of Oral Biology, 26,809-814.
CHAPTER 10
1. INTRODUCTION
Growth of an organism in biological terms begins with the fertilised ovum, followed
by the 'process of self-multiplication of living substance' (Malina 1975). After the
initial stages in the proliferation of cells, the generalised cells get involved in the
process of differentiation and specialisation as different functional units. This later
activity continues and the initial embryo begins to develop the species specific
bodily characteristics leading towards the birth and maturity of the individual. This
process is usually referred as the developmental process. Tanner (1990) does not refer
to such distinctly defined views (meanings). He is comfortable with the use of the
word growth in a wider perspective. In general, the word growth has been
interchangeably used for both in most of the writings on the subject. In a very recent
discussion in tracing the evolution of human growth Bogin (1 999) appears to have a
similar approach. Expressions in Bogin's (1999) book are no exception although the
glossary does suggest separate meanings for the two words. The meanings are
overlapping and in reality the two go side by side without specific boundaries,
except where a particular case (a feature) in point is under discussion.
Irrespective of the fine nuances between the two terms as some would prefer to
have for the sake of definition or clarity on specific issues, the fact of the matter is
that the initial genetic components and the environment, pre-natal as well as post-
natal, both contribute towards the attainment of final bodily attributes of an
individual. It, obviously, involves the interaction between the genetic and the
environment which, in tum, controls and directs the whole process of development.
Beyond the normal control of mechanisms of growth and development, factors like
nutrition, diseases, temporal infections, cultural constraints, geographical factors
affecting food supplies, periodic or long term, and many other unknown, heavily
influence the net results in the developmental outcome.
Potentials of genetic components under normal circumstances have been
substantiated by the twin studies, but a potential may not be fully reached because of
malnutrition or childhood diseases of a particular nature, or some unknown factors in
operation.
Some positive aspects of the influence of nutritional and health environments,
however, have been reported with regards to the stature in various human
123
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 123-128.
© 2001 Kluwer Academic Publishers.
124 CHAPTER 10
is financially weak, occasional insufficient supply of staple food items like, wheat,
gram, lentils, rice and potato, etc. puts them in a disadvantageous situation. Meat
proteins hardly form 0.5% of the diet, but they do eat meat once in a blue-moon.
Families Band C are vegetarians.
Table 1 Stature of individuals (in cm) in successive generations in the three families
There was hardly any difference in the dietary patterns and items in earlier
generations which grew up before the 1960's. Until then the diet was mostly based
126 CHAPTER 10
on cereals, lentils, occasional use of potatoes and some dairy products - mainly milk
and some butter and buttermilk in average and well-to-do families in the village.
Among the dairy products, only buttermilk was occasionally available to family A.
The oldest members in the families A and B only grew up during World War I,
and the dry seasons as well as the depression of the early 1930's. The second
generation grew up during the 1930's and witnessed the societal phases in the 1940's
and after. Since all the families (including C) were directly dependent on village
economy (farming and labour) they were more or less influenced along similar lines.
Therefore, I presume that food supply and the physical environment had similar
influence on the growing children and sub-adults in the population. With this
background, I proceed to look into the patterns of stature attainment in the three
families stated above.
The three families scrutinised here are patrilocal and patrilineal. As a result, the
information is available only for the males in the family and their female siblings.
The information on the family lines of the daughters of the families married outside
and away from the village could not be obtained. The spouses of the males within
the village do not willingly participate in such observations. Hence their stature was
assessed in comparison and in relation to the stature of another member in the family
wherever possible. This problem was more acute with newly arrived brides in the
family. This situation gives only a limited reliability on the recorded figures
presented here. The number of individual families is small, although the information
on four successive generations is made available. In family C, the fifth generation is
also available. Also, all members of each generation in each family are not alive.
Family A is represented by the members in the third and fourth generations, family
B by fourth and fifth generations and family C by second, third and fourth
generations.
The readings for the stature of individuals were available in inches, and are
presented in Table 1 after conversion to centimetre.
Family A and C represent four generations, while B goes to the fifth generation.
The individuals in each family are designated serially, from the start to end. The
male and female siblings are identified; the respective spouses are listed separately.
The cross mark(s) in this column indicate the non-availability of the information.
The data for the children of the female siblings in a family were not available except
in family A because the sibling resided in the village, along with her brothers.
3. DISCUSSION
The limited reliability of the figures (for stature) and the number of individuals
observed, is a drawback over which there was, and is, no control. And possibly no
standardised statistical procedures would have yielded any more reliable results than
the systematic observations on the trend in the increase and decrease of stature in
successive generations in each family.
Family A indicates a gradual decrease in stature of the males in each generation.
In family B, there are ups and downs. Here, the male 3 (son) drops down
considerably in stature against his father with 172.7 cm; even 5.1 cm shorter than
his mother with 162.6 cm. The third generation sons (5 and 7) gain back in stature
(165.1 and 170.2 cm, respectively). The younger male (7) gains back a considerably
higher stature of 170.2 cm than his older brother who is only 165.1 cm. In
RIDDLES IN HUMAN DEVELOPMENT PATTERNS 127
generation 4, the two sons (9 and 11) have a slight gain in stature over their father
(individual 5). However, in the same generation 4, the son (13) drops in stature by
3.8 cm as against his taller father (7). Thus, we notice here first a considerable drop
in the second generation, and a gain in the third generation. The generation 4 males,
however, do not indicate a considerable variation as in generation 2 and 3. The
relative increase of stature in the third generation may be partially accounted for
better conditions of living (in terms of availability of food resources), but this factor
of diet does not seem to have any influence on individuals in generation 4, where
there is somewhat better dietary environment. A question may be raised here as to
what happened to the genetic factors commencing in the tallest male individual (1)
in the first generation and expressed only in one individual in the third generation?
At the same time how does one explain the extreme loss in the second generation
male. Extreme environmental conditions have not been reported. On the other hand
members in the fourth generation have not reported any obvious gain back with
somewhat still better dietary resources.
Family C commences with average stature in the population, but a considerable
drop in stature similar to family B in the second generation is noticed. The third
generation gains in stature over their father (154.9 cm) but certainly not over their
mother (167.6 cm) except in one case (6). The males in generation 4, maintain more
or less (within 1.3 cm) the stature of the third generation, but slightly shorter than
their respective fathers. Undoubtedly, the nutritional environment for the generations
3 and 4 was better during the growth period than the generation 2. The dietary
environmental factors, here, may be interpreted to have lent their influence in stature.
But the introduction of a taller mother in the second generation must not be
overlooked. Her stature of 167.6 cm is considerable for a woman in that area, as well
as in the community. However, the lone female child (5) remains considerably
smaller as compared to her mother, though within the normal range of stature for
females; the allele(s) for the tall stature from the mother may have worked (if at all)
along with the improved nutritional environment for males only.
In family A where there is gradual decline in stature, individuals (as spouses) of
obviously higher or lower stature have not been added, and the nutritional
environmental factors have remained almost similar.
For specifically short stature, Family A is marked by individual 2, a female,
family B by individual 3, a male and family C by individual 3, also a male. The
spouses of these shorter individuals were taller and the fathers of the shorter males in
B and C were taller as well.
An overview of the data, thus, does not reveal a clear mode of inheritance and the
expression of stature attainment. If the dietary environment is to be considered as a
major factor, the consumption of some milk in family Band C in their third and
fourth generations may have played a role to attain the potential, which, in the
circumstances of the families, may be considered a responsible factor, since the
'stature is a dynamic phenotypic trait, one that is very responsive to the quality of
the environment for growth' (Bogin 1999). In reference to his discussion on the
inter-generational effect hypothesis, Bogin further comments that 'a child's height is
a historical record of both the individual and his or her parents'. The historical
record in the three families does point to the fact that average, or taller individuals
did exist among the parents of the individuals, who were either very short or average
and taller. Here, then, the arguments may be put forward in favour of the hereditary
128 CHAPTER 10
factors. Where the quality of environment for growth was better, the trait responded
in favour of taller genetic potential. But in the extreme cases of shortness, as in the
family Band C, the adverse environment for growth has not been reported. These
individuals do not even fall in between the stature of their parents. The question for
such a drop begs an explanation. Besides, family A registers a continuous downward
trend in more or less similar environment for growth.
Thus, the riddle in the expression of stature influenced by genetic or
environmental factors still remains unclear. It appears imperative that more refined
tools to examine the mode of working of the genetic factors, in a given environment
which influences it, are needed.
4. REFERENCES
Bogin, B., 1999, Evolutionary Perspectives on Human Growth. In Annual Review of Anthropology,
volume 28 edited by W.H. Durham (Palo Alto: Annual Reviews) pp. 109-310.
Bogin, B., 1999, Patterns of Human Growth (Cambridge: Cambridge University Press).
Ganguly, P., 1979, Progressive Decline in Stature in India: a study of sixty population groups. In
Physiological and Morphological Adaptation and Evolution, edited by M.A. Stini (Hague: Mouton),
pp.315-317.
Kromeyer - Hauschild, K., and Jaeger, U., 2000, Growth Studies in Jena, Germany: changes in sitting
height, biacromial and bicristal breadth in the past decenniums. American Journal of Human
Biology, 12,646-654.
Malina, R.M., 1975, Growth and Development: The First Twenty Years in Man (Minneapolis: Burgess).
Meredith, H.V., 1963, Changes in stature and body weight of North American boys during the last
eighty years. In Advances in Child Development and Behaviour, volume 1, edited by L.P. Lipsitt
and C.C. Speker (New York: Academic) pp. 63-114.
Meredith, H.V., 1976, Findings from Asia, Australia, Europe, and North America on secular changes in
mean height of children, youths, and young adults. American Journal of Physical Anthropology, 44,
315-326.
Stein, P.L., and Rowe, B.M., 2000, Physical Anthropology (New York: McGraw-Hili).
Takaishi, M., 1995, Growth standards for Japanese children - an overview with special reference to
secular changes in growth. In Essays on Auxology, edited by R. Hauspie, G. Lindgren and F.
Falkner (Welwyn Garden City, U.K. Castlemead) pp. 302-311.
Tanner, J.M., 1990, Foetus into Man: Physical Growth from Conception to Maturity (Cambridge,
Massachusetts: Harvard University Press).
Tobias, P.V., 1985, The negative secular trend. Journal of Human Evolution, 14,347-356.
CHAPTER 11
M. L6PEZ-BLANCO
Fundaci6n Bengoa. Caracas-Venezuela
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 129-135.
© 2001 Kluwer Academic Publishers.
130 CHAPTER 11
130
l
21
c
<U • Energy
E
~ 120
• Protein
.:; * Iron
0"'
~
B 110
00
c
...
:i:)
0
(.)
<.)
100
'"00
0)
.fl
c0)
90
~
0)
p...
80
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
~ri~
Figure 1. Levels of sufficiency in energy, protein and iron availability 1989 - 1999.
Instituto Nacional de Nutrici6n - Universidad de Los Andes. Roja de Balance de Alimentos
(1989 - 1999). Data adjusted by Magaly Bracho and Magaly Rangel (INN)
130
• 1994
• 1995
~ 1996
120
o 1997
~ 1998
s:: 110 ~ 1999
~ 100
90
80
Stratum III Stratum IV Stratum V
Figure 2. Energy intake adequacy according to social strata in Caracas 1994 - 1999.
Fundacredesa - Estudio sobre condiciones de vida 1994 - 1999.
l32 CHAPTER 11
16
15
~fj 14
~
13
12
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
Period
Figure 3. Low weight-for-age prevalences in 0 - 2 year olds in Venezuela: Instituto
Nacional de Nutricon (INN) - Sistema de Vigilancia Alimentar{a Nutricional (SISVAN)
1990 - 1999. cut-off point: percentile 10 of NCHS/OMS
40
35
'E
cu
~
cu
30
Q..
25
20 90 91 92 93 94 95 96 97 98 90 91 92 93 94 95 96 97 98 99
2 - 6 years 7 - 14 years
Figure 4. Undernutrition prevalencest in children 2 - 14 years of age in Venezuela 1990 -
1999. Instituto Nacional de Nutricion (INN) - Sistema de Vigilancia Alimentar{a y
Nutricional (SISVAN) 1990 - 1999. t including low weight-far-height and low height-for-
age. Cut-off points: percentiles 10 and 3 of NCHS/OMS, respectively.
NUTRITION IN VENEZUELA 133
15
14
13
f 12
~ 11
10
9
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
Period
Figure 5. Low birth weight prevalence in Caracas 1989 - 1998: Instituto Nacional de
Nutrici6n (INN). Sistema de Vigilancia Alimentar{a Nutricional (SISVAN) 1989-/998.
A preliminary report indicates that the positive secular trend in height and weight
in the SISV AN data shows a tendency to reverse. However, data from an ongoing
study of Fundacredesa, clearly shows that the secular trend (at selected ages of
9.11,13 and 15 years of age in girls and boys) is still present, although higher in
height than in weight and most evident at ages 9 and 11 (Mendez de Perez, Landaeta
de Jimenez, Ledezma and Ortega, 1999). Also, the magnitude of this trend, analysed
between 1984 and 1995, (0.5-1 cm/decade) is lower than the one reported previously
134 CHAPTER 11
between 1936 and the 1980's, probably due to the socio-economic crisis of this
decade (LOpez-Blanco et at. 1989).
5. REFERENCES
Brundtland, G. H., 1980, Height, weight and menarcheal age of Oslo school children during the last 60
years. Annals of Human Biology, 7, 307-322.
Fundacredesa, 1998a, Indicadores de Condiciones de Vida 1996-1997, Area Metropolitana de Caracas.
(Caracas: Fundacredesa)
Fundacredesa, 1998b, Estudio Impacto del Enriquecimiento de Harinas con Hierro y Vitamina A en la
Poblaci6n Venezolana. 1998 (Caracas: Fundacredesa)
Fundacredesa, 1999, Indicadores de Condiciones de Vida 1998, Area Metropolitana de Caracas
(Caracas: Fundacredesa).
Fundacredesa, 2000, Indicadores de Condiciones de Vida 1999, Area Metropolitana de Caracas.
(Caracas: Fundacredesa)
Instituto Nacional de Nutrici6n (INN)/Sistema de Vigilancia Alimentaria y Nutricional (SISV AN),
1998, Boletin Informativo 1994-91.
Instituto Nacional de Nutrici6n (INN)/Sistema de Vigilancia Alimentaria y Nutricional (SISV AN),
1999, Boletin Informativo.
Instituto Nacional de Nutrici6n (INN) and Universidad de los Andes (ULA) 1989-1999, Hojas de
Balance de Alimentos. (Merida: Talleres Gnificos Universitarios ULA).
Lindgren, G., 1976, Height, weight and menarche in Swedish urban school children in relation to
socioeconomic and regional factors. Annals of Human Biology, 3, 501-528.
L6pez de Blanco, M., Landaeta de Jimenez, M., Izaguirre de Espinoza, I., Macias de Tomei, c., 1995,
Crecimiento Fisico y Maduraci6n. In Estudio Nacional de Crecimiento y Desarrollo Humano de la
Republica de Venezuela: Proyecto Venezuela. (Vol 2), edited by H. Mendez Castellano. (Caracas:
Escuela Tecnica Popular Don Bosco), pp 695-773.
L6pez de Blanco. M., 1999, EI pediatra ante el Hambre Oculta. Anales Venezolanos de Nutrici6n,
12,129-136.
L6pez-Blanco. M., 1995, Growth as a Mirror of Conditions of a Developing Society: The Case of
Venezuela. In Essays on Auxology. Presented to James Mourilyan Tanner, edited by R Hauspie, G
Lindgren, Falkner (London: Castlemead Publications), pp 313-321.
L6pez-Blanco. M., Landaeta-limenez. M., and Mendez-Castellano. H., 1989, Secular trend in height
and weight: Carabobo, Venezuela. 1978-1987. In Auxology 88: Perspectives in the Science of
Growth and Development, edited by. J.M. Tanner. (London: Smith Gordon), pp 207-210.
NUTRITION IN VENFZUELA 135
Mendez-Castellano. H., L6pez-Blanco. M., Mendez, M., Fossi, M., Landaeta-Jimenez. M., and Bosch,
M., 1990, The social impact on child growth and development in Venezuela. In Malnutrition and
the Infant Brain: Neurology and Neurobiology. (New York: Wiley Liss Inc), pp 269-284.
Mendez de Perez, B., Landaeta de Jimenez. M., Ledezma, T. and Ortega de Mancera. A., 1999,
Tendencia secular en peso y talla entre 1984 y 1995 en niiios y j6venes venezolanos. Anales
Venezolanos de Nutrici6n, 12,117-122.
Oficinal Central de Estadfstica e Informatica (OCEI) 1993-1998, Mapa de la Pobreza: Basado en los
resultados del XII censo de poblaci6n y vivienda 1990.
Paez-Celis, J., 1995, Marco situacional de la Poblaci6n Venezuela. In Proyecto Venezuela 1981-
1987.(Vol. I), edited by H. Mendez Castellano (Caracas: Fundacredesa). pp 17-43
Tanner, J. M., 1986, Growth as a mirror of the condition of society: secular trends and class distinctions.
In Human Growth, a Multidisciplinary Review, edited by A. Demirjian and M. Brault-Dubuc.
(London and Philadelphia: Taylor and Francis), pp. 3-24
van Wieringen, 1. C., 1986, Secular Growth Changes. In Human Growth, a Comprehensive Treatise,
(Vol 3), edited by F. Falkner and J. M. Tanner. (New York and London: Plenum Press, 2nd edn),
pp. 307-331.
CHAPTER 12
P.B. EVELETH
1. INTRODUCTION
Changes in body size, fatness, rate of growth, and timing of maturation have been
occurring over the past 150 years. We also are witnessing a decline in mortality, an
increase in life expectancy and a large increase in the proportion of the population
that is elderly in both industrialised and developing countries.
We assume that the secular increase in size and earlier maturation is a result of
improved environmental conditions within a population. These have led to improved
health status of the population as a whole. Similarly, secular decrease in size or
delayed maturation may come about because of negative conditions. But the reasons
remain only partially uncovered. Quite assuredly, many factors help children to more
closely approach their full genetic potential. Among these are improved nutrition,
more widespread health and medical care, higher education level of mother (maybe
the father, too), better housing and living conditions, better transportation, clean
water, improved sanitation, immunisation programs and population mobility, both
geographically to urban areas and socially upward. This can be linked, both in time,
and in conditions with the Industrial Revolution in Europe and the United States
when there were tremendous changes in society, some beneficial and some
detrimental. Early in the Industrial Revolution in Europe during the late 18th and
early 19th centuries, children were employed in factories and mines and working
families were densely packed in cities. Conditions were such that children were short,
birth weight was low and mortality was high (Tanner 1981, 1986). Only later, in the
early 20th century, when advances in science and technology improved standards of
living of ordinary people, did height begin to rise (Fogel 2000). Today many
developing countries are experiencing industrialisation with the shift from an
agrarian economy to a manufacturing and commercial one accompanied by a shift in
population from the country to the cities. Some already are witnessing an increase in
children's height and weight, and a shift to the right in distribution of Body Mass
Index (BMI).
137
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 137-145.
© 2001 Kluwer Academic Publishers.
138 CHAPTER 12
30 • Asiatic
·········+·····~·········+···········i··· • White
• Hispanic
25 )( Black
.....
§ 20 ..........~ ............ +- .......... + ........... ~ ....... ·····r··~·········· ················r ············r··········
u
~
~ : : : : : :
15
. . . . ~. . . . ~. mN··m
10
197919801981198219831984198519861987198819891990
Period
There are a number of adoption studies, all showing catch-up growth in height
and weight, the greater the stunting, the greater the acceleration of growth. In a study
of Indian girls adopted in Sweden at 1 month to 11 years of age, Proos (1993)
reported considerable improvement in growth. However, she noted shorter final adult
height in some cases in spite of rapid growth because menarche began sometimes as
early as 7-9 years of age, thereby limiting much additional height growth.
Some health workers in developing countries do believe that differences among
populations in child growth will disappear (Habicht et al. 1974). That indeed is the
basis for the World Health Organisation's (WHO) proposition to construct a new
international growth reference (Garza and de Onis 1999). This would be a single
standard for the world's children and replace the current WHO reference which is
based on the US National Center for Health Statistics data. It has been proposed that
seven geographic sites be selected from North and South America, Sub-Saharan
Africa, Europe, and East, South and West Asia. In order for new-borns to be selected
the parents must follow WHO feeding recommendations, i.e. exclusive breast feeding
for the first 4 to 6 months and partial breast-feeding until 12 months of age.
3. TALLER
It may be surprising to some but it appears that humans are not taller today than
they have ever been in history or prehistory.
According to long bone measurements of some fossil men reported by Styne and
McHenry (1993) and shown in Table 1 early European and West African males were
7 cm taller than modern Homo sapiens.
Table 1. Evolution of height (in em) in the Hominids (modified from Styne and McHenry
1993)
Records of US Civil War soldiers have revealed that their average stature was
considerably lower than that of American males 100 years earlier, showing a secular
decrease in stature. Fogel (2000) proposes that the largest contributor to this decrease
was over-rapid urbanisation, combined with poor sanitation and diet. In the twentieth
century average heights in the US and in some European countries began to rise, as
shown in Table 2, probably due to improved nutrition as the food supply increased
(Fogel 1986, 1993, 2000).
140 CHAPTER 12
At other times and in other places conditions have been right for optimal growth
in height. It is quite likely that tallness has been cyclical throughout human history.
Table 2. Estimated average heights (in em) of men who reached maturity between 1750
and 1875 (modified from Fogel 1986, 1993)
However, the class difference in height apparently has disappeared now in Sweden
(Lindgren 1976) and perhaps in Norway (Brundtland et aI., 1980). Not only is the
assumption made that a positive secular trend reflects an improvement in social
conditions, the recommendation is that child growth be used to monitor conditions
in the population at large (Tanner 1981). It is assumed that this is due to
environmental effects but recently Bielicki (2000) presented one hypothesis (among
others) of a genetic influence.
There are public health and societal implications involving an increase in height.
Tall individuals may be healthier. Fogel et al. (1982) have shown that large size is
highly correlated with increased work capacity and labour productivity and also with
lower mortality. Waaler (1984), working with extensive Norwegian data, also
demonstrated reduction in mortality with greater height. All adults in Norway (except
SECULAR TRENDS IN GROWTH AND DEVELOPMENT 141
Oslo) were measured between 1963 and 1975. Males and females from 40 to 59
years, separated into five year age groups, showed decreasing relative mortality by
increasing height except for the very tall (> 185 cm for men and > 170 cm for
women).
Taller people have lower risk of myocardial infarction than shorter people
(Palmer et at. 1990, Barker et at. 1989). Tall women have fewer low birth weight
babies. On the negative side, Micozzi (1987) and other researchers have reported an
increase in cancer incidence at some sites and in cancer mortality with increasing
height (Table 4).
Table 4. Correlation of age-specific mean stature with age-adjusted breast cancer mortality
(adapted from Micozzi 1987)
4. FATTERIHEAVIER
The secular trend also is towards heavier and fatter people. There are clear public
health implications of the secular trend toward more overweight and obesity. The US
national surveys show that average weight, triceps and subscapular skinfold have
been rising in children and adolescents over the past 20 years.
Table 5. Estimated prevalence of obesity in US black and White children and adolescents:
based on triceps skinfolds ~ 85th percentiles HES 11 & III (from Malina 1993, as adapted
from Gortmaker et at. 1987)
Males Females
Age group/survey Black White Black White
6-11 years
RES cycle II 8.1 19.5 9.5 18.6
NHANES I 9.6 25.0 20.0 22.8
NHANESII 16.6 31.5 20.9 26.0
12 - 17 years
RES cycle II 7.5 16.7 12.8 16.6
NHANESI 8.9 17.6 20.4 24.5
NHANESII 12.7 19.5 25.1 25.6
Table 5 shows the increasing prevalence of obesity in black and white children from
the Health Examination Survey (HES), collected 1963-70, the National Health and
142 CHAPTER 12
Nutrition Examination Surveys (NHANES I), collected 1971-4, and NHANES II,
collected 1976-80, (Gortmaker et al. 1987, Kuczmarski 1993).
The trend is not limited to the US. In fact, Price et at. (1993) have suggested
there may be a world-wide epidemic of obesity. Studies documenting the
unfavourable health outcomes of overweight and obesity in children and adults are
increasing as rapidly as the problem itself. The well-recognised links to diabetes,
cardiovascular disease, gall bladder disease and some site-specific cancers leaves little
question as to the increased risks involved in an individual being heavier or fatter.
Furthermore, the population prevalence of these chronic diseases is increasing as
well. There are social implications of obesity since in Europe and the US, the
increase in prevalence is found among the lower, not upper, socio-economic groups.
5. PROPORTIONS
Body proportions change with secular increase in stature. In Japan and in China the
increase in stature was due almost entirely to an increase in leg length. Among
Japanese children relative leg length has increased in the past four decades (Ali 2000).
The major part of the increase in height of Norwegian recruits between 1921 and
1962 was due to increase in leg length (Udjus 1964). Similar results were seen in
Harvard students from the late 1900's to 1930 (Himes 1979). Previously we had
thought body proportions were immutable, but it appears that is not so. The reason
may be that the greatest influence of environmental improvement occurs during early
childhood when the legs are the fastest growing part of the body. Thus, the
relationship of sitting height to stature is altered.
6. SEXUAL MATURATION
Secular change also has been occurring in rate of maturation. This is shown by age
of menarche which has been getting earlier during the past 100 years, by some three
to four months per decade, in most European countries. The trend seems to be
slowing down in most developed countries or even stopped; but one cannot predict
when these changes will cease. Many environmental factors influence the mean age
of menarche. Some that have not been considered until recently are hormone
disrupters, especially man-made oestrogen mimics, such as DDT, polychlorinated
biphenyls (PCB), and organophosphates. It is thought that these may cause earlier
sexual maturation (Colborn et at. 1996, Holden 1997). Since menarche is brought
on by the heightened activity of the sex hormones one must consider the possible
influence of natural phyto-oestrogens and oestrogen mimics in the environment.
7. NATURAL RESOURCES
The secular trend will influence a nation's resources and economic development. On
the one hand, taller people require more calories for maintenance, while on the other
hand, taller people have higher productivity and less disease (WHO 1995).
Fogel in his Nobel Prize address (Fogel 1993) summarised some of his work on
secular trend in energy intake, productivity and economic development.
Significantly, he believes that the food supply in Europe during the eighteenth
SECUlAR TRENDS IN GROWTH AND DEVELOPMENT 143
century would not have been sufficient to maintain tall people and have much energy
left over to sustain work. Hence stature remained low as height data from Europe
show (see Table 2). The trend in increasing height came as nutrition improved
among all classes of people.
Other studies on work capacity and BMI in India and some African countries
indicate lower work capacity with lower BMI (Shetty and James 1993). In a sample
of llOO Rwanda women those with low BMI spent more time resting and less time
doing productive work (Figure 2) .
.£50
>-
45
.~ 40 ~ ...
~
'0
...
•
~
BMI < 17.1
BMI 17. 1 - 17.5
CIl
Q) 35 ... [Ill] BMI 17.6 - 18.6
;:-
~
:: 30
0 BMI 18.7 - 23.8
~
t
~
25
15
~ 20
.....
(\3
~ 15
8
.';:: 10
.....
c~
(.) 5
I-<
~
A... 0
laying down maintenance heavy work
Level of activity
Figure 2. Effect of BMl on physical activities of Rwandese women, 1982 (after Fram;ois
P., FA 0, 1990, unpublished data)
It is important to document the ongoing secular trend both positive and negative.
The world is changing at a rapid pace and many of these changes may influence child
growth, such as, the revolution in Yugoslavia, violence in Israel and Indonesia,
hurricanes in Mexico, floods in Bangladesh, AIDS in Africa, East Asia, Europe and
the Americas, newly emerging diseases, development of new drugs and access to
them, drought in Africa and Australia. Health workers, economists and politicians
now understand the value in recording anthropometric measurements of well-defined
samples at regular intervals to use in the evaluation of past policies and current
conditions. It is hoped that this will influence the formulating of national policies in
the future.
144 CHAPTER 12
REFERENCES
Ali A., 2000, Secular changes in relative leg length in post-war Japan. American Journal of Human
Biology, 12,405-416.
Barker, D.J.P., Osmund, c., Golding, J., Kuh, D., and Wadsworth, M.E.I., 1989, Growth in utero, blood
pressure in childhood and adult life and mortality from cardiovascular disease. British Medical
Journal Journal, 298, 564-567.
Bielicki T., 1986, Physical growth as a measure of economic well-being of populations: The twentieth
century. In Human Growth - A Comprehensive Treatise, 2nd ed., vol 3, edited by F. Falkner and
1.M. Tanner (Plenum Press, New York), pp 283-305.
Bielicki T., and Szklarska, A., 2000, Are social class differences in stature partly genetic? A hypothesis
revisited. American Journal of Human Biology, 12,97-101.
Brundtland, G.H., Leistol, K., Walloe, L., 1980, Height, weight and menarcheal age of Olso
schoolchildren during the last 60 years. Annals of Human Biology, 7, 307-322.
Colborn, T., Dumanoski, D., and Myers, J.P., 1996, Our Stolen Future (Penguin USA, New York).
Fogel, R.W., 1986, Physical growth as a measure of the economic well-being of a popUlation: The
eighteenth and nineteenth centuries. In Human Growth - A Comprehensive Treatise, 2nd ed., vol
3, edited by F. Falkner and J.M. Tanner (Plenum Press, New York), pp 263-281.
Fogel, RW., 1993, Economic growth, population theory, and physiology: The bearing of long-term
processes on the making of economic policy. Presentation as the Prize Lecture in Economic
Sciences in Memory of Alfred Nobel, December 9, 1993.
Fogel, RW., 2000, The Fourth Great Awakening & the Future of Egalitarianism (University of Chicago
Press, Chicago).
Fogel, RW., Engerman, S.L., and Trussell, J., 1982, Exploring the uses of data on height: the analysis of
long-term trends in nutrition, labor welfare and labor productivity. Social Science History, 6, 401.
Garza, C., and de Onis, M., 1999, The selection of reference data in the assessment of growth: the new
World Health Organization reference. In Human Growth in Context edited by F.E. Johnston, B.
Zemel, P.B. Eveleth (Smith-Gordon, London).
Gortmaker, S.L., Dietz, W.H. jr., Sobal, A.M., Wehler, C.A., 1987, Increasing pediatric obesity in the
United States. American Journal of Diseases of Children, 141, 535-540
Greulich, W.W., 1957, A comparison of the physical growth and development of American-born and
native Japanese children. American Journal of Physical Anthropology 15,489-515.
Habicht, J.-P., Martorell, R, Yarborough, C., Malina, RM., Klein, RE., 1974, Height and weight
standards for preschool children. How relevant are ethnic differences in growth potential? Lancet
i,611-615.
Himes, J.H., 1979, Secular changes in body proportions and composition. Monographs of the Society for
Child Development, 44, 28-58, 103-120.
Holden, C., 1997, Early puberty getting more common. Science 276, 537.
Kuczmarski, RJ., 1993, Trends in body composition for infants and children in the U.S. Critical
Reviews in Food Science and Nutrition 33 (4/5), 375-387.
Lindgren, G., 1976, Height, weight and menarche in Swedish urban school children in relation to socio-
economic and regional factors. Annals of Human Biology, 3, 501-528.
Malina RM., 1993, Ethnic variation in the prevalence of obesity in North American children and youth.
Critical Reviews in Food Science and Nutrition, 33, 389-396.
Micozzi, M., 1985, Nutrition, body size and breast cancer. Yearbook of Physical Anthropology, 28,
175-206.
Palmer, J.R., Rosenberg L., and Shapiro, S., 1990, Stature and the risk of myocardial infarction in
women. American Journal of Epidemiology, 132,27-32.
Price, A.B., Charles, M.A., Petit, D.J., Knowler, W.C., 1993, Obesity in Pima Indians: Large increases
among post-World War II birth cohorts. American Journal of Physical Anthropology, 92, 473-480.
Proos, L.A., 1993, Anthropometry in adolescence-secular trends, adoption ethnic and environmental
differences. Hormone Research 39 (suppl 3), 18-24.
Shetty, P.S., and James, W.P.T., 1993, Body Mass Index: A Measure of Chronic Energy Deficiency in
Adults. FAO, Rome.
Styne, D.M., and McHenry, H., 1993, The evolution of stature in humans. Hormone Research 39 (suppl
3),3-6.
Tanner, J.M., 1981, A History of the Study of Human Growth (Cambridge University Press,
Cambridge).
Tanner, J.M., 1986, Growth as a mirror of the conditions of society: Secular trends and class
distinctions. In Human Growth: A Multi disciplinary Review edited by A. Demirjian (Taylor &
Francis, London).
Udjus, L.G., 1964, Anthropometrical Changes in Norwegian Men in the 20th Century
(Universitetsforlaget, Olso).
SECULAR TRENDS IN GROWTH AND DEVELOPMENT 145
Waaler, H.T., 1984, Height, weight and mortality: the Norwegian experience. Acta Medica
Scandinavia, suppl. 679, 1-5.
WHO, 1995, Physical Status: The Use and Interpretation of Anthropometry. WHO Technical Report
Series 854, WHO, Geneva, p. 346.
Yip, R., Scanlon, K., Trowbridge, F., 1992, Improving growth status of Asian refugee children in the
United States. Journal of the American Medical Association, 267, 937-940.
CHAPTER 13
A.F.ROCHE,B.TO~
1. INTRODUCTION
We are delighted by the opportunity to contribute to this commemorative volume
that honours a leader responsible for major contributions to our communal
knowledge of child growth. Professor Das is one of relatively few investigators who
have initiated and conducted longterm serial studies of child growth; he may be the
only one among those few who made all the measurements himself. This
extraordinary data collection effort has already led to important papers dealing with
growth status, growth patterns, sibling comparisons and sexual dimorphism. His
data continue to be a great resource for future studies of secular trends and many other
aspects of child growth.
The term "secular trends" refers to differences among groups within a population
that are explained totally or largely by differences in birth dates. If the range of birth
dates is less than about ten years, any observed growth differences that are related to
birth dates are more likely to be due to sampling variations than to secular trends.
Exceptions to this statement may occur, however, when the environmental
influences on growth change very rapidly, as can occur during wars and famines.
Over the last two centuries, secular trends in child growth have generally taken the
form of increases; that is, children at a particular age are larger and more mature now
than in the past. Such secular trends are commonly referred to as "positive", but this
terminology is somewhat undesirable today because secular increases in weight,
skinfold thicknesses and body mass index, within populations that were already well-
nourished, have negative effects on health. Furthermore, in such populations, secular
increases in stature may not be associated with health benefits. Secular increases of
considerable magnitude have occurred in Europe and the United States, but the
occurrence and the size of secular trends are uncertain for many countries due to the
lack of national surveys and sampling differences among studies. Secular increases
are particularly well documented for the Czech Republic, Japan, the Netherlands, and
the United States where sampling and anthropometric methods have remained fixed
among repeated national surveys (Fredriks et al. 2000, Freedman et al. 2000, Kimura
1984, Lhotska et al. 1993, Troiano et al. 1995).
147
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 147-157.
© 2001 Kluwer Academic Publishers.
148 CHAPTER 13
trends may have influenced the applicability of the results. Since 1920, prenatal and
neonatal care have improved in developed countries, and diseases during infancy and
childhood have become less common, but weight and body mass index have
increased over the last two decades, especially in adolescence and adulthood, despite
greater awareness of the health risks associated with overweight and obesity. Infants
born now with small birth weights, and adolescents with large weights or large body
mass index values, may not be subject to the health consequences indicated by these
retrospective data. There have been reductions in mortality rates for infants with
small birth weights and for adults with cardiovascular diseases.
Most longterm serial studies enrolled a single cohort. Well-known examples are
the Berkeley and Harvard studies in the United States (Jones et at. 1971, Stuart
1939). The participants in the Berkeley study were born in 1928 or 1929, and those
in the Harvard study were born from 1930 to 1939. Consequently, their growth may
have been affected by the Great Depression. Other notable examples are the Belgian,
English, French, Swedish and Swiss studies that began in 1954 and were co-
ordinated by the Centre International de I'Enfance (1980). Due to a lack of random
sampling, these studies cannot provide satisfactory cross-sectional reference data for
representative groups. Despite sampling limitations, cross-sectional reference data
have been developed from some of them (Karlberg et at. 1976, Prader et at. 1989,
Sempe et at. 1979). The utility of these reference data has been reduced, however, by
secular trends in the populations to which they were meant to be applied (Demoulin
1998, Lindgren 1998, Vercauteren et al. 1984). This statement should not be
interpreted as a criticism of these studies. They were designed to investigate patterns
of change and of relationships among variables; not to provide cross-sectional
reference data for representative populations. Still, they are useful and, in some
cases, unique sources of data.
A few longterm serial growth studies have enrolled participants over a large
number of years. The best, and perhaps only, examples are the United States studies
at the Child Research Council in Denver, Colorado, and the Fels Longitudinal Study
in Dayton, Ohio (McCammon 1970, Roche 1992). The Denver study enrolled 334
participants who were born from 1915 to 1966 and examined until 1968. In the Fels
Longitudinal Study, about 15 infants were enrolled at birth each year from 1929 to
the present except from 1975 through 1982 when financial constraints delayed
enrolments until the children were aged one to seven years. The Figure shows the
design of the Fels Longitudinal Study. The oblique lines indicating the nature of the
serial data for each annual group illustrate the opportunities for analysing age-to-age
correlations and growth patterns using these data. Both the Denver study and Fels
Longitudinal Study have been sources of reference data for growth status, despite the
admonitions stated in this paper, in part because they were considered better than
other data available at the time (Hamill et at. 1977, McCammon 1970).
The applicability of the data from the Denver study and the Fels Longitudinal
Study could be reduced by secular trends in the general population and/or within the
studies. Secular trends in the general population would reduce the applicability of
reference data for growth status, and increments or patterns of change, especially for
groups enrolled many years ago. There is a lack of information about secular trends
in the entire United States population prior to 1963. Since then, secular trends for
size at an age have been small except for weight and body mass index which have
increased markedly since 1980 for children aged six years and older (Troiano et al.
150 CHAPTER 13
1995). The absence of secular trends in stature, at least partly, reflects migration
from countries where statures are smaller.
70
60
50
~ 40
~
~
<: 30
20
10
o ~~~~~h~1G~~~~~~~~~~~~~~44~44~~4~
Secular trends within the Fels Longitudinal Study may need to be considered
before data from the many annual groups are pooled for analyses. These trends have
been estimated by regressions on year of birth and (year of birth)2 and by
comparisons among groups defined by decade of birth. In the Fels Longitudinal
Study data, recumbent length and stature at an age became larger until the early
1950s and then became smaller, but these changes were slight and not significant
(Byard and Roche 1984). Similarly, Kouchi et al. (1985a, b) showed that parameters
of functions fitted to serial data for weight and recumbent length during infancy
changed only slightly among participants grouped by decade of year of birth. Bock
and Sykes (1989) confirmed these findings, but demonstrated inter-generational
increases in stature at an age in like-sex comparisons. Apparently, these intra-
SECULAR TRENDS AND GROwrn STUDIES 151
Prahl-Anderson and her colleagues believed that this design reduced possible
confounding due to examination effects and secular trends.
Examination effects, which are sometimes called testing effects or learning
effects, are changes in the recorded data due to the repetition of examinations.
Examination effects occur in behavioural variables and auditory thresholds and
probably explain why blood pressures are lower in serial studies than in cross-
sectional studies (Guo et al. 1998, Kallman and Jarvik 1959, Roche et al. 1983).
Examination effects are of limited interest in the present context because they do not
occur in anthropometric data nor, with the exception of blood pressure, in risk
factors for later disease that may be related to size during childhood. In theory, the
Nijmegen design could help address possible secular differences among cohorts, but
the cohorts in the Nijmegen design were born no more than six years apart.
C;onsequently, any inter-cohort differences are more likely to be due to sampling
variations than to secular trends. In fact, inter-cohort differences for anthropometric
variables were present for head circumference only; this increased significantly with
birth date in boys and in girls. It is also of interest that the anthropometric status
values from the Nijmegen Study exceeded those from a concurrent national survey in
the Netherlands (de Wijn 1976). This reinforces the view that status values from
serial studies should not be relied upon as reference data.
A design similar to that of the Nijmegen Study is logistically difficult, and
substantial financial support is needed from the commencement. Most longterm
serial growth studies have begun with modest support that increased gradually
(Roche et al. 1999). The Fels Longitudinal Study began simply and inexpensively
with a group of six infants, but became more complex and more expensive due to
enlargement of the sample, increases in the number of variables recorded, and the
need for more complex analyses as serial data accumulated.
An alternative design, sometimes called "shingling" or "laddering", is an
extension of the Nijmegen plan. In concept, if one were to enrol participants at
birth, 4, 8, 12 and 16 years and measure each participant serially for four years, one
would obtain a complete description of growth from birth to 20 years. This goal
might be met if sampling were consistent across groups, if the groups did not differ
in enrolment response rates and missed examinations, and if, on final analysis, the
pieces of the total curve join without disjunctions due to differences in levels or rates
of growth. A major disadvantage of this approach is that it requires large
expenditures from the beginning, and it is difficult to obtain such financial support.
There are also limitations to what can be learned. The combination of data from
cohorts differing in birth dates can provide reference percentiles for increments, but
they do not allow analyses of tracking or risk of a later outcome over periods longer
than those for which each cohort was measured. Clearly, decisions about the design
of a longterm serial study must be linked to the central aims of the study.
trends can both strengthen and weaken a new serial investigation and the need to take
these into account during the design phase. While the primary factors responsible for
secular trends are generally understood (e.g., improvements in nutrition and control
of disease), secular trends remain dynamic processes that vary across populations.
Conducting serial studies of growth in developing countries offers an opportunity to
identify specific agents of secular change and to assess the magnitude of their effects.
In so doing, the varied processes of secular change may be more clearly understood.
A pilot study of the growth and development of children from the Jirel ethnic
group in eastern Nepal has recently been initiated. The Jiri Growth Study (Towne et
al. 1999,2000) is conducted in collaboration with the Jiri Helminth Project, a study
of genetic factors predisposing individuals to helminthic infection (Williams-
Blangero et al. 1999). The goal of the Jiri Growth Study is to quantify genetic and
environmental factors influencing variation among children in their growth and
development. The Jirel population is ideal for such a study, because most of the
approximately 4,000 Jirels are members of one very large extended pedigree.
Chronic infection with parasitic disease has particularly negative effects on
children's health. Gastrointestinal diseases, especially diarrhoea, malabsorption of
nutrients, and anaemia, lead to deficits in growth and development. In some cases,
the immune system is compromised, opening the door to other infections. Overall,
at any point during childhood, approximately 50% of Jirel children are infected with
at least one species of helminth, either roundworm, hookworm, or whipworm. Pilot
data from a sample of 431 children aged 3 to 18 years showed that helminthic
infections have deleterious consequences for the growth of Jirel children. Boys and
girls infected with roundworms or hookworms were significantly shorter than non-
infected children, and they were smaller in virtually all body size measures.
Besides various diseases, there are other environmental factors detrimental to
growth and development that are commonly placed in sociocultural, socio-economic,
or socio-demographic categories. These categories may serve as useful proxies for
measures of nutrition and hygiene, which are difficult to measure directly. Lack of
sanitation and poor hygienic practices contribute to the spread of infectious disease,
particularly geohelminthic gastrointestinal infections. Sociocultural factors including
availability of potable water, indoor plumbing, use of latrines, and food handling
practices that pertain to the spread of such diseases. Sociocultural and socio-
economic surveys of households participating in the Jiri Helminth Project show that
household income is strongly associated with helminthic infections (Subedi et al.
2000, Williams-Blangero et al. 1998). Household income is associated with a cluster
of characteristics that are risk factors for transmission of helminths. Among these are
presence or absence of a household latrine, and whether or not each individual in the
household wears shoes. Other characteristics associated with household income are
number of rooms in the house, and the type of house construction. Thus, household
income appears to serve as a useful overall measure of household cleanliness and
sanitation, which are important in the transmission of helminthic infection to
children who stay in and around the house most of the day. For roundworm,
hookworm, and whipworm, analysed separately, there were highly significant non-
random associations of infection status with household income, such that infected
Jirel children came predominately from low income households, while uninfected
children came predominately from high income households. These findings were
154 CHAPTER 13
even more pronounced when infection status was defined as being infected with more
than one species of helminth.
Clearly the growth of Jirel children is negatively impacted by parasitic infections.
It is likewise similarly impacted by other factors associated with poor household
environment (e.g., caloric and, especially, protein deficiencies). Measures of growth
and development, however, have significant genetic determinants. Ideally, both
environmental exposure and familial relationship data are analysed simultaneously to
evaluate the relative contributions of genes and environment to growth and
development. To evaluate the roles of roundworm infection and household
environment on the stature of Jirel children, the heritability of stature was estimated
in genetic models with and without inclusion of roundworm infection status (non-
infected vs. infected) and/or household income (low vs. high) as covariates. The best
statistical genetic model of the stature of Jirel children contained significant covariate
effects of both roundworm infection status and household income on the stature of
boys, such that infected boys are shorter than non-infected boys, and boys from
higher income households are taller than boys from lower income households.
Interestingly, these covariate effects were not significant for girls. Nonetheless, the
h2 of stature with both sexes considered was very high at approximately 90%.
Hopefully, the Jirel population, like the rest of the developing world, will see
decreases in the prevalence of disease and increases in personal wealth over the next
few decades. Such changes will contribute to increasing numbers of children being
healthier and growing larger than at present. These secular trends will not complicate
the objectives of the Jiri Growth Study, rather, by its very structure, the Jiri Growth
Study will offer an opportunity to quantify specific aspects of the environment that
result in secular trends in growth.
4. CONCLUSION
The preceding comments might lead a reader to conclude that longterm serial growth
studies are too problematic to either initiate or continue. This would be a gross error!
The major purposes of such studies are to describe growth patterns, predict future
growth, analyse the relative importance of determinants of growth and development,
and evaluate the significance of changes during childhood and adulthood. Without
such past studies, we would lack accurate descriptions of infant growth, the
pubescent growth spurt, and changes in the body mass index, and we would not have
methods to predict adult stature and assess skeletal maturity, to mention just a few of
their products. Further, such studies are needed in different populations, both normal
and abnormal, and should include measures other than those that describe size. In
particular, data should be collected for studying determinants of growth (e.g., familial
and genetic data), and for factors that increase the risk of diseases in adulthood. There
are clear guidelines for the conduct of these studies, and the necessary statistical
methods are readily available. More, not fewer, longterm serial growth studies are
needed.
SECULAR TRENDS AND GROWTH STUDIES 155
5. REFERENCES
Barker, D.J.P., Hales, e.N., Fall, C.H.D., Osmond, C., Phipps, K, and Clark, P.M.S., 1993, Type 2 (non-
insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to
reduced fetal growth. Diabetologia, 36, 62-67.
Barker, DJ.P., Winter, P.D., Osmond, C., Margetts, P., and Simmonds, SJ. 1989, Weight in infancy and
death from ischaemic heart disease. Lancet, 2, 577-580.
Bayley, N., and Pinneau, S.R, 1952. Tables for predicting adult height from skeletal age: revised for
use with the Greulich-Pyle hand standards. Journal of Pediatrics, 40, 423-441. (Erratum corrected
in Journal of Pediatrics, 41, 371.)
Billewicz, W.z., and McGregor, LA., 1982, A birth-to-maturity longitudinal study of heights and
weights in two West Africian (Gambian) villages, 1951-1975. Annals of Human Biology, 9, 309-
320.
Bock, RD., and Sykes, RC., 1989, Evidence for continuing secular increase in height within families in
the United States. American Journal of Human Biology, I, 143-148.
Bodzsar, E.B., and Susanne, e., 1998. Secular Growth Changes In Europe, Budapest, Eotvos University
Press.
Byard, PJ., and Roche, A.F., 1984, Secular trend for recumbent length and stature in the Fels
Longitudinal Growth Study. In Human Growth and Development, edited by 1. Borms, R Hauspie,
A. Sand, C. Susanne, and M. Hebbelinck (New York: Plenum Press), 209-214.
Centre International de I'Enfance, 1980, Growth and Development of the Child. Courrier, 30, 1-260.
Demoulin, F., 1998. Secular trend in France. In Secular Growth Changes in Europe, Budapest, edited by
E. Bodzsar and e. Susanne (Budapest: EOtvos University Press), 109-134.
Fredriks, A.M., van Buuren, S., Burgmeijer, RJ.F., Meulmeester J.F., Beuker R.I., Brugman E., Roede
MJ., Verloove-Vanhorick S.P., and Wit, J.-M., 2000, Continuing positive secular growth change in
the Netherlands 1955-1997. Pediatric Research, 47,316-323.
Freedman, D.S., Khan, L.K., Serdula, M.K., Srinkvasan, S.R, and Berenson, G.S., 2000, Secular trends
in height among children during 2 decades: The Bogalusa Heart Study. Archives of Pediatric and
Adolescent Medicine, 154, 155-161.
Guo, S.S., Chi, E., Wisemandle, W., Chumlea, W.C., Roche, A.F., and Siervogel, RM., 1998, Serial
changes in blood pressure from childhood into young adulthood for females in relation to body mass
index and maturational age. American Journal of Human Biology, 10, 589-598.
Guo, S.S., Chumlea, W.e., Roche, A.F., Gardner, J.D., and Siervogel, R.M., 1994, The predictive value
of childhood body mass index values for overweight at age 35 y. American Journal of Clinical
Nutrition, 59, 810-819.
Guo, S.S., Chumlea, W.e., Roche, A.F., and Siervogel, RM., 1997, Age- and maturity- related changes
in body composition during adolescence into adulthood: The Fels Longitudinal Study. International
Journal of Obesity, 21,1167-1175.
Hamill, P.V.V., Drizd, T.A., Johnson, C.L., Reed, RB., Roche, A.F., 1977, NCHS Growth Curves for
Children. Birth-18 years, United States. Vital and Health Statistics, National Center For Health
Statistics, Series II, No. 165, pp. 74,1977.
Hauspie, RC., Das, S.R, Preece, M.A., Tanner, J.M., 1982, Degree of resemblance of the pattern of
growth among sibs in families of West Bengal (India). Annals of Human Biology, 9, 171-174.
Jenss, RM., and Bayley, N., 1937, A mathematical method for studying the growth of a child. Human
Biology, 9, 556-563.
Jones, M.C., Bayley, N., MacFarlane, J.W., Honzik, M.P., 1971. The Course Of Human Development
(Waltham, MA: Xerox College Publishing).
Kallman, FJ., and Jarvik, L.F., 1959, Individual differences in constitution and genetic background. In
Handbook of Aging and the Individual, edited by J.E. Birren (Chicago: University Press), 216-263.
Karlberg, P., Taranger, J., Engstrom, I., Karlberg, J., Landstrom, T., Lichtenstein, H., Lindstrom, B.,
and Svennberg-Redegen, I., 1976, Physical growth from birth to 16 years and longitudinal outcome
of the study during the same age period. The Somatic Development of Children in a Swedish Urban
Community. A Prospective Longitudinal Study, Acta Paediatrica Scandinavica, Supp1258, 7-70.
Kimura, K., 1984, Studies of growth and development in Japan. Yearbook of Physical Anthropology,
27, 179-214.
Kouchi, M., Mukherjee, D., and Roche, A.F., 1985a, Curve fitting for growth in weight during infancy
with relationships to adult status, and familial associations of the estimated parameters. Human
Biology, 57,245-265.
Kouchi, M., Roche, A.F., and MukheIjee, D., 1985b, Growth in recumbent length during infancy with
relationships to adult status and familial associations of the estimated parameters. Human Biology,
57,449-472.
156 CHAPTER 13
Lhotska, L., Blaha, P., Vignerova, Roth, Z., and Prokopec, M.V.V., 1993, Celostatnf Antropologicky
Vyzkum Deti A Mladeze 1991 (Ceske Zeme): Antropometricke charakteristiky (Praha: National
Institute of Public Health).
Lindgren, G., 1998, Secular growth changes in Sweden. In Secular Growth Changes in Europe, edited
by E. Bodzsar and C. Susanne (Budapest: E5Mls University Press), 319-333.
McCammon, R.W., 1970, Human Growth and Development (Springfield, IL: Charles C Thomas).
Must, A., Jacques, P.F., Dallal, G.E., Bajema, e.1., and Dietz, W.H., 1992, Long-term morbidity and
mortality of overweight adolescents. A follow-up of the Harvard Growth Study of 1922 to 1935,
New England Journal of Medicine, 327, 1350-1355.
Nieto, F.J., Szklo, M., and Comstock, G.W., 1992, Childhood weight and growth rate as predictors of
adult mortality. American Journal of Epidemiology, 136,201-213.
Prader, A., Largo, R.H., Molinari, L., and Issler, C., 1989, Physical growth of Swiss children from birth
to 20 years of age. First Zurich Longitudinal Study of Growth and Development. Helvetica
Paediatrica Acta, S52, 1-125
Prahl-Anderson, B., 1979, The conduct of the Nymegen Growth Study. In A Mixed-Longitudinal
Interdisciplinary Study of Growth and Development edited by B. Prahl-Anderson, e.J. Kowalski,
and P. Heydendael (New York: Academic Press), 61-70.
Preece, M.A., and Baines, MJ., 1978, A new family of mathematical models describing the human
growth curve. Annals of Human Biology, 5, 1-24.
Prentice, A.M., Cole, T.J., and Foord, F.A., Lamb, W.H., and Whitehead, R.G., 1987, Increased
birthweight after prenatal dietary supplementation of rural African women. American Journal of
Clinical Nutrition, 46, 912-925.
Roche, A.F., 1979, Secular Trends in Human Growth, Maturation and Development. Monographs of the
Society for Research in Child Development (edited), 44, 1-120.
Roche, A.F., 1992, Growth, Maturation and Body Composition: The Fels Longitudinal Study 1929-1991
(Cambridge, UK: Cambridge University Press), 345.
Roche, A.F., Guo, S.S., and Towne, B., 1999, The establishment and maintenance of longitudinal growth
studies: Experiences from Melbourne and Fels. In Human Growth in Context, edited by F. Johnston,
B. Zemel, and P.B. Eveleth. (London, UK: Smith-Gordon), 101-120.
Roche, A.F., Mukherjee, D., Chumlea, W.e., and Champney, T.F., 1983, Examination effects in
audiometric testing of children. Scandinavian Audiology, 12,251-256.
Roche, A.F., Mukherjee, D., Guo, S., and Moore, W.M., 1986. Head circumference reference data:
birth to 18 years. Pediatrics, 79, 706-712.
Roche, A.F., Roberts, J., and Hamill, P.V.V., 1978, Skeletal maturity of youths 12-17 years: Racial,
geographic area, and socioeconomic differentials, United States. Vital and Health Statistics,
National Center for Health Statistics, Series 11, No. 167,98.
Roche, A.F., Wainer, H., and Thissen, D., 1975, Skeletal Maturity: The knee joint as a biological
indicator (Plenum, New York), 374.
Rolland-Cachera, M.-F., Bellisle, F., and Sempe, M., 1989, The prediction in boys and girls of the
weightlheight2 index and various skinfold measurements in adults: A two-decade follow-up study.
International Journal of Obesity, 13, 305-311.
Rowland, M.G.M., Paul, A., Prentice, A.M., MUlier, E., Hutton, M., Barrell, R.A.E., and Whitehead,
R.G., 1981, Seasonality and the growth of infants in a Gambian village. In Seasonal Dimensions to
Rural Poverty, edited by R. Chambers, R. Longhurst, and A. Pacey (London: Francis Pinter), 164-
175.
Sempe, M., Pedron, G., and Roy-Pernot, M.P., 1979, Auxologie methode et sequences (Paris:
Laboratorie TMraplix).
Siervogel, R.M., Roche, A.F., and Guo, S., Mukherjee, D., and Chumlea, W.C., 1991, Patterns of
change in weightlstature 2 from 2 to 18 years: Findings from long-term serial data for children in the
Fels Longitudinal Growth Study. International Journal of Obesity, 15,479-485.
Stuart, H.C., 1939, Studies from the Center for Research in Child Development, School of Public Health,
Harvard University. I. The Center, the group under observation, sources of information, and studies
in progress. Monographs of the Society for Research in Child Development, 4, No.1.
Subedi, J., Subedi, S., Sidky, H., Singh, R., Blangero, J., and Williams-Blangero, S., 2000, Health and
health care in liri. Contributions to Nepalese Studies. April-May, in press.
Susanne, C., 1980, Developmental genetics of man. In Human Physical Growth and Maturation:
Methodologies and Factors, edited by F.E. Johnston, A.F. Roche, and C. Susanne (New York:
Plenum),221-242.
Thomson, J., 1955, Observations on weight gain in infants. Archives of Diseases in Childhood, 30, 322-
327
Towne, B., Blangero, J., Williams-Blangero, S., 1999, Growth and body composition of Jirel children.
American Journal of Human Biology, 11,131.
Towne, B., Blangero, J., Williams-Blangero, S., 2000, Genetic and environmental Influences on the
growth status of Nepali children. American Journal of Human Biology, 12,278.
SECULAR TRENDS AND GROWTII STUDIES 157
Troiano, R.P., Flegel, K.M., Kuczmarski, R.I., Campbell, S.M., and Johnson, C.L., 1995, Prevalence and
trends in overweight for children and adolescents: The National Health and Nutrition Examination
Surveys, 1963-1991. Archives of Pediatrics and Adolescent Medicine, 149, 1085-1091.
Vercauteren, M., Susanne, c., and Orban, R., 1984, Secular evolution in Brussels between 1960 and
1980. In Human Growth and Development, edited by J. Borms, R. Hauspie, A. Sand, C. Susanne,
and M. Hebbelinck (New York: Plenum Press), 215-223.
Wijn, J.F., de, 1976, Zesde orientering omtrent de voedingstoestand van achtjarige schoolkinderen in
Nederland (1973174). Tijdschrift Kindergeneesk, 44, 18-32.
Williams-Blangero, S., Subedi, J., Upadhayay, R.P., Maural, D.R., Khanda, K., lirel, S., Robinson, E.S.,
and Blangero, J., 1998, Attitude towards helminthic infection in the Jirel population of eastern
Nepal. Social Science and Medicine, 47, 371-379.
Williams-Blangero, S., Subedi, 1., Upadhayay, R.P., Maural, D., Rai, D.R., Jha, B., Robinson, E.S., and
Blangero, J., 1999, Genetic analysis of susceptibility to infection with Ascaris lumbricoides.
American Journal of Tropical Medicine and Hygiene, 60, 921-926.
Wright, C.M., Matthews, J.N.S., Waterston, A., and Aynsley-Green, A., 1994, What is a normal rate of
weight gain in infancy? Acta Paediatrica, 83, 351-356.
CHAPTER 14
M.HENNEBERG
1. INTRODUCTION
159
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 159-167.
© 2001 Kluwer Academic Publishers.
160 CHAPTER 14
of changing specific elements of the environment while health policies may ensure
optimal conditions for child growth by setting standards for those specific elements
in addition to the obviously needed general socio-economic improvement.
Figure 1. Secular trends in body height of White and Black South African males and
Khoisan males. Data from Tobias (1975, 1985), Henneberg and van den Berg (1990) and
Louw and Henneberg (1997). Note that slopes of the regression lines do not differ
significantly
162 CHAPTER 14
All slopes were, however, significantly different from zero indicating positive,
but very slow increase of stature. During a comparable period, stature of Dutch
people, from whom the majority of South African Whites are descended, increased at
a rate of about 15 mm1decade (van Wieringen 1986).
It is interesting to note that adult South Australian Aborigines recently surveyed
by ourselves (Pretty et at. 1998) display lack of strong secular increase of stature
similar to oppressed South African Blacks (Figure 2). This occurs despite
documented substantial welfare provided to Aborigines by the government during the
second half of the 20th century. Among some Australian children of European
ancestry a secular increase comparable to those reported for Europe and the US has
been found (Loesch et at. 2000), but this acceleration of physical development did
not translate into substantial increases of adult stature. Data on adult statures of
Australians, though, are scarce. Stature of 5000 adult (18 - 65 years) females
surveyed in 1926 was 1611 mm (Lancaster 1957) while the stature of another 4327
women of comparable age surveyed in 1995 by the Australian Bureau of Statistics
(1995) was 1624 mm. This indicates a rate of about 2 mm/decade. A study of
Australian military men indicated the secular increase of stature at a rate of about 4
mm/decade (Soar 1999). These rates are similar to the weak increase found among
South Africans and to the increases among Aboriginal Australians. The socio-
economic history of Australia is similar to that of Europe and the United States, but
the increase in adult stature seems to be less pronounced.
1900
e-- Aborigines
1850 o Blacks o
o
1800 o
~
o
1750 o o
E
E o~
1700 o -u co
19
....
A
..<::
01)
0 &ctJ 0 i oD
'Q) o 0 I:lll DCO@
::r:: 1650 o 0 0
00
1600
1550
1500
1890 1910 1930 1950 1970 1990
Year
Figure 2. Comparison of secular trends in body height of South African Blacks (sample
means like in Fig. 1) and South Australian Aborigines (individual heights by birthdate
adjusted for loss of stature with age, own anthropometric observations, Pretty et at. 1998)
Soviet-style policies. Yet the secular trend in stature was strong, exceeding 10
mm/decade (Piontek 1971, Wolanski 1978, Bielicki and Welon 1982). Traditionally,
Polish medical students came from the upper socio-economic strata of the society. In
a sample of 300 students observed by ourselves in 1983 (Henneberg et al. 1985)
64% of parents had full university education. Similar situation obtained in the 1930
study of 153 medical students by Wrzosek (1931). This contrasts with about 5% of
university-educated individuals in the country. Post-war policies clearly favoured the
upliftment of people of lower socio-economic backgrounds at the expense of
economic well-being of the upper and middle classes. Despite this alteration of the
rate of change of socio-economic conditions for various socio-economic groups, the
rate of secular trend among the medical students (12 mm/decade) was the same as
among conscripts representing the total population (12 mm/decade) (Figure 3).
Portugal has been considered one of the least socio-economically developed
countries of Western Europe. Sobral (1990) presented data on body height changes in
four provinces of Portugal (1930-1980). For each province data were analysed
separately for rural and for urban conscripts. In all provinces, and at all dates
examined, rural people were shorter than their urban counterparts, yet in all cases
trends had similar rates of about 10 mm/decade (Figure 4). Parallelism, once again,
though socio-economic conditions and their temporal changes different from those of
Poland.
1900
S 1750
S
.._r 1700
,..t:i
b/)
.(j)
::r:: 1650
1600
1550
1500
1910 1930 1950 1970 1990
Year
Figure 3. Secular trends of body height in Polish males. High socia-economic status is
represented by medical students, data from Henneberg and van den Berg (1990).
164 CHAPTER 14
1700
e-- Setubal, urban
1690 .• _.& .. - Setubal, rural
e-- Beja, urban
1680
- - - - E}- - - - Beja, rural
1620
1930 1940 1950 1960 1970 1980
Year
3. DISCUSSION
The positive secular trend of stature was still going strong in the 1970's and 1980's
in some economically advanced countries, while it has apparently halted in others
(Roche 1979, Malina 1990). Were it true that secular trend is a result of consecutive
generations reaching increasingly greater proportions of their ideal "genetic
potential", privileged groups in various societies would have to halt their trends
SECULAR TRENDS INDICATE SPECIFIC FACTORS 165
earlier than underprivileged ones, who should continue until they "caught up" with
their wealthier compatriots. This certainly was not the case neither in South Africa
nor in Poland. Moreover, the greatest differences in stature between the poorest and
the richest segments of same national populations do not exceed one standard
deviation, i.e. about 60 - 70 mm (Bielicki and Welon 1982, Rosenbaum et al. 1985,
Henneberg and Louw, 1998), while positive secular trends, where they occurred
strongly, shifted stature by as much as 120 - 200 mm (van Wieringen 1986,
Spurgeon et al. 1994). This, coupled with high heritability of the variation in
stature, argues strongly against secular trends being a simple ecosensitive response
to improving general living conditions. In addition, the magnitude of secular trends
in populations experiencing similar changes in socio-economic status varies
considerably from 2 - 6 mm per decade among Australians and South African Whites
to 10 - 15 mm among Europeans.
Secular trends must be caused by specific factors varying from population to
population rather than by the general improvement in living conditions. General
socio-economic improvement may play some role, but it can hardly cause changes
exceeding 70 mm. We cannot yet name specific factors causing greater increases in
height, but certain criteria that they must meet can be specified. This may help to
direct our search for specific factors. Since statural increases in some instances exceed
the range of ecosensitivity, the causative factor must either affect genes determining
stature, or act as a substitute for one of the relevant gene products. If the second is
true then this factor must act early in ontogeny, sometime during the foetal period,
infancy or early childhood, as these are the periods in which most of the body size
determination seems to occur (Bogin 1988, Proos 1993, Delemarrevandewaal 1993,
Henneberg and Louw 1990, 1993). These criteria are met by changing exposure to
pathogens, natural or contained in vaccines, which produce immune responses that
may be biochemically linked to determinants of stature, and by foods which contain
chemical substances influencing regulators of growth. For instance, beef and poultry
may contain additional animal growth hormones used to obtain higher yield on
commercial farms (Moishezon-Blank 1991). Ingestion of dairy products containing
similar natural or artificial compounds is another possibility. Bigger is not always
better, and continued secular increases of body height may be manifestations of the
introduction of undesirable substances into human bodies.
Henneberg, M., 1999, Variation in hominid body size: Do we know how big our ancestors were?
Perspectives in Human Biology, 4, 65-81.
Dettwyler, K.A., 1992, Nutritional status of adults in rural Mali, American Journal of Physical
Anthropology, 88, 309-321.
Eiben, O.G., 1996, Selected data on growth and physical fitness of healthy Hungarian boys and girls in
our changing world, Collegium Anthropologicum, 20, 73-81.
Falconer, D.S., 1980, Introduction to Quantitative Genetics (Edinburgh: Oliver and Boyd).
Frayer, D., 1984, Biological and cultural change in the European Late Pleistocene and Early Holocene.
In The Origins of Modem Humans: A World Survey of the Fossil Evidence edited by F.H. Smith
and F Spencer (New York: Alan R Liss), pp 211-250.
Gerver, W.J.M., Debruin, R, and Drayer, N.M., 1994, A persisting secular trend for body
measurements in Dutch children - The Oosterwolde II study, Acta Paediatrica, 83, 812-814.
Gutierrez-Muniz, J.A., Berdasco-Gomez, A., Esquivel-Lauzurique, M., Jimenez-Hernandez, J.M.,
Mesa-Payan, D., Posada-Lima, E., Romero-Del-Sol, J.M., Quesdada, M.R., and Silva-Ayzaguer,
L.C., 1994, The 1982 Cuban national growth and development study: Secular changes in weigbt and
height, Acta Medica Auxologica, 25, 115-127.
Hauspie, R.C., Vercauteren, M., and Susanne. C .. , 1996, Secular changes in growth, Hormone
Research, 45: (Suppl. 2), 8-17.
Hauspie, R.C., Vercauteren, M., and Susanne, C., 1997, Secular changes in growth and maturation: an
update, Acta Paediatrica, Supplement, 423, 20-27
Henneberg, M., 1976, Reproductive possibilities and estimations of the biological dynamics of earlier
human populations, Journal of Human Evolution, 5,41-48
Henneberg, M., 1988, Decrease of human skull size in the Holocene, Human Biology 60, 395-405.
Henneberg, M., 1992, Continuing human evolution: brains, bodies, and the role of variability,
Transactions of the Royal Society of South Africa 48, 159-182.
Henneberg, M., Budnik, A., Pezacka, M., and Puch, A.E., 1985, Head size, body size, and intelligence:
intraspecific correlations in Homo sapiens sapiens. Homo, 36, 207-218.
Henneberg, M., and Louw, GJ., 1990, Height and weight differences among South African Urban
schoolchildren born in various months of the year, American Journal of Human Biology, 2, 227-233
Henneberg, M., and Louw, G.J., 1993, Further studies on the month-of-birth effect on body size: Rural
schoolchildren and an animal model, American Journal of Physical Anthropology, 91, 235-244.
Henneberg, M., and Louw, GJ., 1998, Cross-sectional survey of growth of urban and rural "Cape
Coloured" schoolchildren: Anthropometry and functional tests. American Journal of Human
Biology, 10, 73-85.
Henneberg, M., and Steyn, M., 1993, Trends in cranial capacity and cranial index in Subsaharan Africa
during the Holocene, American Journal of Human Biology, 5,473-47.
Henneberg, M, van den Berg, E.R, 1990, Test of socioeconomic causation of secular trend: Stature
change among favored and oppressed South Africans are parallel, American Journal of Physical
Anthropology, 83, 459-465.
Huang, Y.C., and Malina, R.M., 1995, Secular changes in the stature and weight of Taiwanese children,
American Journal of Human Biology, 7,485-496.
Jacobs, K.H., 1985, Climate and the hominid postcranial skeleton in Wilrm and early Holocene Europe.
Current Anthropology, 26, 512-514.
Jamison, P.L., 1990, Secular trends and the pattern of growth in arctic populations, Soc. Sci. Med., 30,
751-759.
Kimura, K., 1967, A consideration of the secular trend in Japanese for height and weight by a graphic
method, American Journal of Physical Anthropology, 27, 89-94.
Lancaster, H.O., 1957, Some anthropometrical values of women in Australia, Medical Journal of
Australia (December 1957), 897-900
Leatherman, T.L., Carey, J.W., Thomas, RB., 1995, Socioeconomic change and patterns of growth in
the Andes, American Journal of Physical Anthropology, 97, 307-321.
Loesch, D.Z., Stokes, K., and Huggins, RM., 2000, Secular trend in body height and weight of
Australian children and adolescents, American Journal of Physical Anthropology, 111,545-556.
Louw, GJ., and Henneberg, M., 1997, Lack of a secular trend in adult stature in South African males
born between 1954 and 1974, Homo, 47,54-61.
Malina, R.M., 1990, Research on secular trends in auxology, Anthropologisher Anzeiger, 48, 209-227.
Mathers, K., and Henneberg, M., 1995, Were we ever that big? Gradual increase in hominid body size
over time, Homo, 46, 141-173.
Moishezon-Blank, N., 1991, Commentary of the possible effect of hormones in food on human growth.
Medical Hypotheses, 38, 273-277.
SECULAR TRENDS INDICATE SPECIFIC FACTORS 167
Mueller, W.H., 1976, Parent-child correlations for stature and weight among school aged children: A
review of24 studies. Human Biology, 48, 379-397.
Oyedeji, G.A., Olamilulo, S.K., Osinaike, A.I., Esimai, V.c., Odunusi, E.O., and Aladekomo, T.A.,
1996, Secular trends in the growth of children aged 0-6 years in a rural Nigerian community,
Annals of Tropical Paediatrics, 16, 11-17.
Pretty, G.L., Henneberg, M., Lambert, K.M., and Pro kopec, M., 1998, Trends in stature in the South
Australian Aboriginal Murraylands, American Journal of Physical Anthropology, 106,505-514.
Piontek, J., 1971, Ocena stanu rozwoju fizycznego poborowych powiat6w Inowroclaw i Tuchola
urodzonych w drugiej poTIowie XIX wieku w por6wnaniu z wsp61czesnll mlodzieza Pomorza i
Kujaw, In Rozw6j i Dorobek Antropologii Polskiej w 25-leciu PRL. Poznaj: UAM, pp 71-75.
Prince, J.M., 1995, Intersection of economics, history and human biology - Secular trends in stature in
nineteenth-century Sioux Indians, Human Biology, 67, 387-406
Prokopec, M., 1984, Secular trends in body size and proportions, and their biological meaning. Studies
in Human Ecology, 6, 37-61
Proos, L.A., 1993, Anthropometry in adolescence - Secular trends, adoption, ethnic and environmental
differences, Hormone Research, 39, 18-24
Rao, K.V., Balakrishna, N., and Shatrugna, V., 1993, Secular trends in height of well-to-do adults and
the associated factors, Man in India, 73, 267-273
Roberts, D.F., 1994, Secular trends in growth and maturation in British girls, American Journal of
Human Biology, 6,13-18
Roche, A.F., 1979, Secular trends in stature, weight and maturation, Monographs of the Society for
Research in Child Development, 44, 2-27
Roberts, D.F., Billewicz, W.Z., and McGregor, LA., 1978, Heritability of stature in a West African
population, Annals of Human Genetics, 42, 15-24
Rosenbaum, S., Skiner, R.K., Knight, LB., and Garrow J.S., 1985, A survey of heights and weights of
adults in Great Britain, 1980. Annals of Human Biology, 12, 115-127
Soar, N., 1999, An examination of the secular trend in Australian male stature as seen in the military
data, Perspectives in Human Biology, 4, 157-166.
Sobral, F., 1990, Secular changes in stature in southern Portugal between 1930 and 1980 according to
conscript data, Human Biology, 62, 491-504
Spurgeon, lH., French, K.E., Giese, W.K., Steele, M.F., Utenko, V.N., Bundzen, P.V., and Rogozkin,
V.A., 1994, Somatic comparisons of urban and rural Russian boys ages 6, 9, and 15 years, living in
St Petersburg and surrounding areas, American Journal of Human Biology, 6,141-151
Susanne, C .. , 1971, Recherche sur la transmission des caracteres mesurables de ('homme, Societe
Royale des Sciences Naturelles de la Belgique: Brussels
Styne, D.M., and McHenry, H., 1993, The evolution of stature in humans, Hormone Research,
39(Suppl. 3), 3-6
Tobias, P.V., 1975, Stature and secular trend among Southern African Negroes and San (Bushmen),
South African Journal of Medical Science 40, 145-164
Tobias, P.V., 1985, The negative secular trend, Journal of Human Evolution, 14,347-356
Tobias, P.V., 1990, Adult stature in southern African Negroes - further evidence on the absence of a
positive secular trend, South African Medical Journal, 78, 97-101
Ulijaszek, S.J., 1993, Evidence for a secular trend in heights and weights of adults in Papua-New
Guinea, Annals of Human Biology, 20, 349-355
van Wieringen, lV., 1986, Secular growth changes. In Human Growth. Vol 3 Methodology Ecological
Genetic and Nutritional Effects on Growth edited by F. Falkner and lM. Tanner (New York:
Plenum Press), pp. 307-331
Weber, G., Seidler, H., Wilfing, H., and Hauser, G., 1995, Secular change in height in Austria - an
effect of population stratification, Annals of Human Biology, 22, 277-288
Wolanski, N., 1978, Secular trend in man: Evidence and factors. Collegium Anthropologicum, 2, 69-86
Wrzosek, A., 1931, 0 stosun~u niekt6rych pomiar6w antropologicznych i typ6w rasowych
morfologicznych do sprawnoEci umyslowej. Przeglad Antropologiczny, 5, 1-16.
CHAPTER 15
1. INTRODUCTION
Though there are many studies comparing anthropometric measurements of children
belonging to low socio- economic group with those of other studies and developed
countries, studies are very few on children belonging to well-to-do communities
(Banik 1982, Vijayaraghavan et al. 1971., Nutritional Foundation of India, Scientific
Report No. 11, 1991, Rao and Sastry 1977, Reed and Stuart 1959, Hamill et al.
1977, Tanner et al. 1966). Earlier studies indicated that students of the well
maintained public schools have better anthropometric measurements than those of
other schools where most of the students belong to lower socio-economic status.
Most of the children are found to have substantial retardation in measurements
reflective of skeleton, muscle and fat. Studies on children of well-to-do families with
good parental education are of interest. Also it is of interest to assess the association,
if any, between grades of body built and scholastic performance. A small-scale study
was carried out on school-age children and adolescents of well-to-do and lower middle
class families in Hyderabad. An attempt was therefore made 1) to study the pattern of
growth in children of different socio-economic status; 2) to assess whether a
combination of anthropometric measurements or indices is better for evaluation of
various grades of under nutrition based on body weight, and 3) to study the relative
importance of nutritional status, socio-economic status and paternal nutritional
status associated with scholastic performance of school-going and adolescent
children.
169
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 169-177.
© 2001 Kluwer Academic Publishers.
170 CHAPTER 15
relative importance of all the discriminant functions were tested with the use of
Mahalanobis D2 and Fisher's F-test (Visweswara Rao 1977, 1996, Visweswara Rao
et al., 1979). Percentage of subjects properly classified and misclassified into grades
of their scholastic performance was also found with the best sets of variables.
differences. The best set of variables related to body weight (%) was traced with the
calculation of the coefficient of detennination, R2 (%) and adjusted R2 (%).
Table 1. Mean and SD of height and weight by age (years) and socia-economic status
(SES): boys
Height, cm Weight, kg
high SES low SES high SES low SES
Age n mean SD n mean SD mean SD mean SD
5 14 107.2 4.92 53 104.4 5.49 16.3 1.47 15.3 l.89
6 29 116.2 5.03 27 108.6 5.04 19.5 3.11 15.9 l.53
7 25 123.1 4.73 33 114.5 5.42 22.1 3.98 17.4 l.78
8 31 126.7 6.27 30 118.8 6.13 23.8 5.38 18.9 2.18
9 28 133.4 5.15 21 125.7 6.27 26.9 4.67 22.0 3.85
10 33 139.1 5.55 33 128.9 6.30 30.0 4.42 23.2 3.18
11 22 144.0 5.87 37 135.0 6.22 33.3 6.43 25.9 3.73
12 33 148.5 6.86 28 137.9 7.81 35.0 6.02 27.5 4.35
13 31 158.4 7.09 22 145.7 8.29 42.2 6.43 33.1 5.94
14 37 159.8 7.48 20 149.5 7.63 44.0 7.83 34.1 4.62
15 38 167.2 6.88 21 158.5 8.01 49.3 8.03 4l.7 6.75
16 23 167.9 6.86 16 162.5 6.41 52.6 8.22 45.2 4.64
17 37 169.8 4.91 18 163.2 6.55 55.8 9.57 48.3 7.25
18 37 170.0 4.83 14 162.4 5.46 55.7 8.28 47.0 7.05
Table 2. Mean and SD of height and weight by age (years) and socio-economic status
(SES): girls
Height, cm Weight, kg
high SES low SES high SES low SES
Age n mean SD n mean SD mean SD mean SD
5 19 110.0 8.35 40 105.0 5.59 16.2 3.30 15.2 l.89
6 20 115.1 6.95 24 110.7 5.39 18.9 4.46 16.7 2.17
7 25 121.5 5.26 30 115.3 4.54 20.6 3.23 18.2 1.80
8 36 129.6 5.67 27 120.2 5.85 24.9 5.96 19.4 2.80
9 29 132.7 6.44 30 122.9 4.83 27.0 5.45 20.9 2.40
10 31 138.9 7.39 27 130.0 6.91 29.2 5.52 23.9 3.14
11 28 145.6 6.94 23 136.3 6.44 35.3 7.75 26.8 4.59
12 30 152.6 5.03 28 14l.0 7.36 41.1 7.64 30.6 5.00
13 38 152.8 6.11 20 145.2 7.84 41.2 6.97 33.2 7.05
14 36 155.9 5.17 25 150.2 6.32 43.3 5.86 37.6 5.09
15 33 156.8 5.55 17 154.5 6.62 47.9 7.68 41.4 6.12
16 29 155.6 5.25 19 151.2 4.86 48.4 8.24 4l.7 5.75
17 33 156.5 5.60 14 154.6 1.31 49.2 6.87 42.9 5.08
18 31 157.5 5.39 9 15l.1 8.41 48.9 5.73 43.4 7.10
GROWTH AND SCHOLASTIC PERFORMANCE 173
Table 3. Mean and SD of arm circumference and fat fold at triceps by age (years)
and socio-economic status (SES): boys
Table 4. Mean and SD of arm circumference and fat fold at triceps by age (years)
and socio-economic status (SES): girls
Table 5. Mean and SD of weight-for-height (%) by age (years) , sex and socio-
economic status (SES)
boys girls
high SES low SES high SES low SES
Age n mean SD n mean SD n mean SD n mean SD
5 14 94.3 3.8 53 91.2 7.9 19 88.5 12.2 40 90.5 7.8
6 29 92.0 9.0 27 88.6 6.3 20 93.3 14.8 24 89.8 7.4
7 25 93.0 10.8 33 85.5 6.6 25 90.5 8.5 30 90.1 5.1
8 31 93.0 11.5 30 86.5 4.2 36 93.0 11.0 27 86.4 8.0
9 28 94.1 11.8 21 88.5 8.7 29 94.5 14.3 30 87.7 7.2
10 33 94.3 9.8 33 87.3 6.9 31 89.6 10.7 27 87.2 6.6
11 22 94.8 10.1 37 87.2 7.7 28 95.2 16.3 23 87.3 10.1
12 33 91.8 9.2 28 87.7 5.8 30 96.3 14.8 28 88.7 9.4
13 31 93.1 12.3 22 90.3 9.0 38 95.9 15.3 20 86.3 9.9
14 37 92.5 13.3 20 87.0 7.2 36 91.2 11.6 25 87.2 7.4
15 38 91.2 11.1 21 89.3 11.4 33 97.1 13.3 17 86.2 9.5
16 23 94.9 11.8 16 90.0 11.0 29 88.3 13.5 19 87.5 12.5
17 37 94.4 14.5 18 89.6 12.1 33 96.3 12.4 14 85.9 10.7
18 37 92.1 13.7 14 84.4 12.3 31 93.5 9.8 9 85.0 12.7
Height with weight-for-height (%) was found to be the best related to weight than
height alone or height with arm or calf circumferences. The coefficient of
determination was closer to 95%. This was found in both sexes and in all age
groups. This was also found in each of the age groups studied by socio-economic
status. Combination of height and weightlheight 2 was also found best related to
weight. The coefficient of determination was around 99%. This was true in boys and
girls for all ages and socio-economic groups. No other combination yielded such a
high coefficient of determination. Similar findings were observed in studies on
infants, preschool children and adults (Visweswara Rao 1976, 1977, Visweswara Rao
et at. 1986, 1991, Raman et at. 1989). In well-to-do children, weight-for-height was
observed to have higher contribution to variations in body weight than height. In
low socio-economic group, height was observed to have higher contributions to
variations in body weight than that of weight-for-height (%). This was found true in
most of the age groups of low socio-economic status except those of beyond 15 or
16 years.
The contribution of height (%) and weightlheight2 individually to variations in
body weight (%) by age and sex were found different. In well-to-do children,
irrespective of age and sex, weight/height 2 was observed to have higher contribution
to variations in body weight (%) than height. In low socio-economic group, height
(%) was observed to have higher contribution to variations in body weight than that
of weightlheight 2 . That was found true in most of the age groups of boys of low
socio-economic group except those of beyond 15 years. Whereas for girls,
contribution of weightlheight 2 was more than height (%) in most of the ages except
5 to 8 years. This variation may be due to variations in grades of development of
skeletal and muscle reflective measurements and prevalence of various forms of
malnutrition of higher or shorter duration.
GROWTH AND SCHOLASTIC PERFORMANCE 175
Of all children, 76.9% of the boys and 79.7 of the girls were with scores of
scholastic performance of 76.0% or above. Sex differences in these scores were not
significant (p > 0.05).
Height, weight, head circumference, child's education, educational status of
parents, occupation of mother, family size and per capita income were different (p <
0.05) between children with poor and good scholastic performance. All these
indicators were better in children of good scholastic performance.
Of all the measurements - weight, fat fold at triceps and calf circumference in
order constituted the best set of measurements related to scholastic performance for
boys. The coefficient of determination of the multiple linear regression models with
these measurements was 7.8 percent. In girls, height and calf circumference in order
were related to scholastic performance. The value of the coefficient of determination
(R2) was 7.5%.
Order of indicators related to scholastic performance were education of mother,
family size and per capita income for boys, whereas in girls, the indicators in order
were family size, type of family, per capita income and education of mother.
Coefficients of determination of the models for boys and girls respectively were 7.1
and 9.0 %.
Of the indicators, maternal weight and weightJheight2 and father's height in that
order were better related to scholastic performance in boys. Coefficient of
determination, R2 (%) was 4.8. In girls, the indicators in order were weightJheight2 ,
weight and height of mothers. Coefficient of determination was 9.3%.
The best set of variables related to scholastic performance was found to be the
levels of education of child and mother and head circumference for boys. The
coefficient of determination was 17.4 percent. In girls, height, education of child,
family size, head circumference and father's weight/height2 were related to scholastic
performance. The coefficient of determination was 17.1 %. Mean values of
anthropometric measurements of height, weight, head and levels of education of
mother, father and child, occupation of mother, family size and per capita income
were different (p < 0.05) between children with poor and good scholastic
performance.
Educational status of mother and child and head circumference of child were found
best for differentiation of good and poor grades of scholastic performance of boys.
Height, arm circumference and weight-for-height of children enhanced discrimination
power. The percentage miscIassification was 33.7(%).
In girls, height and father's weightJheight 2 were found best for differentiation of
variations in scholastic performance. Percentage miscIassification was around 30.0.
Head circumference, education of father and maternal height and weightJheight2
enhanced the improvements in scholastic performance.
176 CHAPTER 15
4. REFERENCES:
Balakrishna, N., 1992, Efficiency of some parametric tests of inference in nutritional anthropometry.
PhD Thesis (Hyderabad: Osmania University). pp 191-209.
Banik, N.D.D., 1982, Semilongitudinal growth evaluation of children from birth to 14 years in different
socio economic groups. Indian Pediatrics, 19,353-369.
Draper, N.R, and Smith, H., 1981, Applied Regression Analysis. Second Edition (New York: John
Wiley & Sons Inc.).
Hamill, P.V.V., Drizd, T.A., Johnson, C.L., Reed, RB. and Roche, A.F., 1977, NCHS growth curves for
children - birth to 18 years. Vital and Health Statistics Series 11. No. 65, DHEW Pub. No. (PHS),
78-1650, USHES-PHS, Hyattsville, M.D.
Indian Council for Medical Research (lCMR), 1971, Growth and physical development of Indian
infants and children - All India. (New Delhi: Indian Council of Medical Research).
Jelliffe, D.B., 1966, Assessment of the nutritional status of the community. WHO monograph series No
53 (Geneva: World Health Organisation).
Jelliffe, D.B., and Jelliffe, E.F.P., 1989, Community nutritional assessment with special reference to less
technically developed countries. (Oxford: Oxford University Press), pp. 13-140, 355-443, 524-550.
Krishna, D., Balakrishna, N., Visweswara Rao, K., and Reddy, PJ., 1996, Selection of predictor
variables in multiple regression: A case study of agreement between procedures. In Biostatistics -
A manual of statistical methods for use in Health, Nutrition and Anthropology, edited by
Visweswara Rao K. New Delhi: Jaypee Brothers Medical Publishers (P) Ltd. pp. 612-621.
Mayachoudhury, and Visweswara Rao, K., 1983, Nutritional status of preschool children and the
associated factors. Indian Journal of Nutrition and Dietetics, 20, 18-29.
Mayachoudhury, and Visweswara Rao, K., 1984, Association of growth status and mental function in
preschool children. Indian Journal of Nutrition and Dietetics, 21,1-18.
Mayachoudhury, Balakrishna, N. and Visweswara Rao, K., 1990, Mental ability of preschool children:
Relative importance of some associated factors. In Statistics in Health and Nutrition, Proc. National
Seminar Oct. 27-29, 1988, edited by Visweswara Rao, K., Radhiah, G., and Narayana, V.
Hyderabad: National Institution of Nutrition. pp. 201-218.
Nutrition Foundation of India, 1991, Growth performance of affluent Indian Children - Growth
standards for Indian children (New Delhi: NFL) Scientific report series 11.
Raman, L., Vasanthi, G., Parvathi, e., Vasumathi, H., Rawal, A., Visweswara Rao, K. and Balakrishna,
N., 1989, Growth and development of infants in urban slums of Hyderabad. Indian Journal of
Nutrition and Dietetics, 26, 196-205.
Rao, D.H., and Sastry, J.G., 1977, Growth Pattern of well-to-do Indian adolescents and young adults.
Indian Journal of Medical Research, 66, 696-706.
Rao, e.R, 1952, Advanced statistical Methods in Biometric Research. (London: John Wiley & Sons
Inc.). pp. 236-272, 435-490.
Reed, RB., and Stuart, H.e., 1959, Pattern of growth in height and weight from birth to 18 years of age.
Pediatrics, 24, 904-921.
Snedecor, G.W., and Cochran, W.G., 1967, Statistical Methods. (Ames: IOWA State University Press),
pp.90-106.
Tanner, J.M., White house, RH., and Takahashi, M., 1966, Standards from birth to maturity for height,
weight, height velocity and weight velocity - British Children. 1965, part II. Archives of Diseases of
Childhood, 41, 613-635.
Vijayaraghavan, K., Singh, D., and Swaminadhan, M.e., 1971, Heights and weights of well-nourished
Indian School Children. Indian Journal of Medical Research, 59, 648-654.
Visweswara Rao, K., 1976, Efficiency of anthropometric indices for the diagnosis of malnutrition. PhD
thesis (Hyderabad: Osmanian University) pp. 60-201.
Visweswara Rao, K., 1977, Analysis of relative importance of factors affecting body weight changes in
preschool children by nutritional status. Bulletin of the International Statistical Institute, 41st session,
New Delhi. 47, 717-721.
Visweswara Rao, K., Reddy, PJ., Narayanan, T.P., and Subhadradevi, V., 1979, Discriminant function
analysis: A case study for evaluation of various forms of Protein energy malnutrition. Indian
Journal of Medical Research, 69, 99-108.
GROWTH AND SCHOLASTIC PERFORMANCE 177
Visweswara Rao, K., Balakrishna, N., and Adinarayana, K., 1986, Critical limits of some
anthropometric measurements and indices for the assessment of nutritional status. Indian journal of
Nutrition and Dietetics, 23, 88-99.
Visweswara Rao, K., Balakrishna, N., Veena, S. and Thimmayamma, B.V.S., 1991, Body mass index in
school age children and adolescents. Indian Journal of Physical Anthropology and Human
Genetics, 17, 113-124.
Visweswara Rao, K., 1996, Biostatistics - A manual of statistical methods for use in Health, Nutrition
and Anthropology, edited by Visweswara Rao, K .. (New Delhi: Jaypee Brothers Medical
Publishers Ltd) pp. 47-69,137-182,237-284,461-506.
WHO, 1995, Physical status: The use and interpretation of anthropometry. Report of WHO expert
committee. (Geneva: WHO) pp. 37-115, 424-452.
CHAPTER 16
1. INTRODUCTION
It has long been thought that physical and mental growth and development in
children proceed to some degree in parallel, early developers in the one sense being
on average early developers also in the other, as seen e.g. in parallel secular trends
(Tanner 1961, Muuss 1970, Husen 1974, Emanuelsson and Svensson 1986), but
there is surprisingly little published evidence that this is the case.
The connection has also been asserted because of the sex differences found in
favour of girls concerning school achievement, mental ability and other overt
behaviour during puberty, since girls on average become physically mature about
two years earlier than boys (Tanner 1962, 1989, Lindgren 1978). However, the sex
differences do not always point in the same direction and the results found for boys
compared to the girls in different subjects and behaviours are contradictory (Anastasi
1958, Maccoby and Jacklin 1974, Kelly 1978).
Also evidence for mental growth spurts analogous to physical growth spurts has
been suggested and actually found (Ljung 1965, Epstein 1974, Lindgren 1979).
More directly, the relationship between physical and mental growth and
development has been studied by comparing mental test scores of children and youth
in relation to their physical growth and maturation (Shuttleworth 1939, Boas 1941,
Nisbet and Illesley 1963, Douglas and Ross 1964, Johannesson 1974, Kohen-Raz
1977, Lindgren 1979, Westin-Lindgren 1982, 1984). The main assumption behind
these studies has been that early maturers would achieve better at school than late
maturers. The results, however, have been inconsistent, and very few studies have
been conducted in the last 15 - 20 years (Tanner 1999). In 1980, however, data for
Swedish schoolchildren were collected in order to investigate further the relation
between the physical/physiological and the mental/cognitive growth and
development during puberty. The purpose of the study was to analyse the pupils'
mental and cognitive ability in relation to their sex, pubertal stage and socio-
economic background. Although the study by now is a bit old, its results have never
been published; the aim of the present paper is to present them.
179
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 179-20l.
© 2001 Kluwer Academic Publishers.
180 CHAPTER 16
2.11nvestigation group
Table 1. Distribution of boys and girls respectively participating in each Grade (Attrition
is given in parenthesis)
The loss of individuals, who were either ill during the investigation period or did not
like to participate in the study, was 24 pupils out of 306 (8%). Thus the study
group consisted of 282 pupils.
Table 2. Distribution of SES-groups in each school as well as for the total group
(percentages are given in brackets)
translation of the test by the present author is given in the Appendix. The test
consists of 20 pairs of nouns (e.g. a car - a bus) and out of five alternatives for each
pair of nouns the subject has to mark the best and the worst similarity respectively
between the nouns of the pair. The alternatives for each pair of nouns were
constructed according to the theory of Piaget; that is the alternative answers were
constructed to give different stages of cognitive thinking from an egocentric to an
abstract stage of thinking (Inhelder and Piaget 1985), as in the example given; you
can crash into them, have wheels, give out exhaust fumes, you can ride in them and
are vehicles (see Figure 1).
Cognitive level Similarities between a car - a bus
Abstract
logic
C = Conceptual < ~ -
.(~ 5. are vehicles
(abstract overriding concepts)
F = Functional 4. you can ride in them
(their function is alike)
PE = Perceptual < I - ""' 3. give out exhaust fumes
(form, size, place)
Figure 1. Similarities between the pair of nouns "a car - a bus" mirroring different levels
of cognitive thinking from an emotional to an abstract level (Translated from Sandgren
1974)
After instruction and some training (see first page of the test in the Appendix)
the children could work with the test without any time limit - thus minimising
stress. This test was easy to administer and the children liked it since no answer was
ever wrong. However, marking the "worst" similarity did not work out well, since
the children in general thought that once the "best" similarity had been chosen the
other four alternatives were all equally worse. Thus choices of the "worst"
similarities are not analysed here.
The tests measuring mental ability and cognitive thinking were administered on the
same occasion and within two weeks of the occasion when the children's pubertal
stages were rated.
Before conducting the study, much work was done to inform the two schools -
including staff, school nurses, parents and children - about the purpose and procedure
of the study. All school personnel, parents and children were informed that they did
not have to participate if they did not want to, and the collected data were to be
treated as strictly confidential. Written consent was asked for and also given by the
parents. On the whole the study was positively received by the school personnel as
well as by the parents and their children. The study was conducted in March 1980.
3. RESULTS
The results will be presented in three parts. First, the results from the pubertal
ratings will be given; secondly, the analysis of the results on the test "Similarities"
measuring cognitive thinking related to sex and pubertal stage; and finally the
results from three-way ANOV A's of the tests measuring mental abilities and
cognitive thinking controlling for sex, pubertal stage and SES-group.
The distributions of pubertal stages in relation to SES-group for boys and girls
respectively are shown in Table 3. (Grades 4, 6 and 8 have been combined in the
Tables, when testing the distributions regarding differences between SES-groups).
No significant differences between the distributions of pubertal stages were found for
SES-groups I, II and III respectively - neither for girls nor for boys (p > 0.05).
Boys Girls
Pubertal SES-groups SES-groups
stage I II III I II III
I 5 15 17 3 11 8
II 6 13 15 3 8 8
III 1 3 6 9 11 11
IV 2 7 8 9 18 8
V 5 7 4 4 17 10
": ~ f •
~ 12
I'::
Q)
10
8
.....
..... ~
i?
Q)
6
> 4
« 2
0
0 2 3 4 5
Cognitive level
16
b 14 ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::G:rad~}f(i4::;I5:ye3.rs[::::::::::::
.0 12
~ 10
II.)
t:IO
8
E!
II.)
6
> 4
« 2
0
o 2 3 4 5
Cognitive level
As can be seen in Figure 2 girls more often than boys consistently tend to
choose answers on the cognitive levels 4 and 5 (the functional and abstract levels)
from Grade 4 up to and including Grade 8. This tendency seems especially marked in
Grade 6. At the same time, of course, boys more often than girls tend to choose
cognitive levels 1, 2 and 3 (the egocentric, particular and perceptual levels) in all
three Grades.
The distributions on levels of cognitive thinking by sex and pubertal stage are
also shown for each Grade in the Appendix: Figures 1 - 3.
Numbers of chosen "abstract" answers were then analysed in relation to sex,
pubertal stage and SES-group for each Grade.
3.3 Mental ability and cognitive level in relation to sex, pubertal stage and SES-
group.
The ANOV A's of the test scores on the two mental ability tests and the test
measuring cognitive level were done after having transformed the raw scores into
stanine-scores. Regarding the test "Similarities" only numbers of answers
concerning the levelS ("abstract"-level) of the cognitive thinking were transformed.
The results from these three-way ANOV A's of test scores related to sex, pubertal
stage and SES-group can be found in the Appendix (Tables 1 - 9). A summary of
the results is given in Table 4 below.
Source of variation
Type III tests
Test Grade Model S P SES SESxP PxSES S x P xSES
Opposites 4
verbal 6
8 ** **
Letter groups 4 ** * **
logical-induc- 6 * **
tive 8 * *
Similarities 4
abstract cog- 6 * *
niti ve level * *
8 **
** p < 0.01 * p < 0.05 - not significant
Regarding the results on the verbal ability test "Opposites" in Grades 4 and 6, there
were no significant differences between either sexes, pubertal stages or SES-groups.
Neither were there any significant interactions between these factors. In Grade 8,
however, the sex difference is significant (p < 0.01) the girls doing better (see also
Figure 3).
186 CHAPTER 16
8,- Verbal ability grade 4 8 iLogic inductive grade 4 8 iCognitive levels grade 4
",6 6 6
~
o
y
~
u
~
'"
gt 4 4
.~
~ 2 2
o --':-----:':c--=----=~---::'-::--'
L' o L' _-':---:'::---::'::,--=-:-L--' o L' _-':---:'L--'-_-'-_'---'
II III IV V
~ 2 2
r
8 8 ognitive levels grade 8
Verbal ability grade 8 Logk ;"d~d" :
",6
~
u
'"gt
(;)
(OJ
l'l
6
4 4
A
'§
~ (')
2 2
Figure 3. Average scores for boys and girls on the tests measuring verbal ability, logical
inductive ability and cognitive level in relation to pubertal stage in Grade 4, 6, and 8
(0: Boys, .: Girls)
MENTAL ABILITY AND COGNITNE THINKING 187
On the logical inductive test "Letter groups" there were in Grade 4 significant
differences (p > 0.05) between children of different pubertal stages in favour of the
early maturing ones (in this case mainly the girls). At the same time there were
significant differences between SES-groups (p < 0.01). Children from SES-groups I
and II performed better than those in SES-group III.
In Grade 6 there was a significant interaction between sex and pubertal stage (p <
0.01). This meant that there were in this grade (age 12 - l3 years) no significant
differences between girls of different pubertal stages, but there were for boys - the
early maturing boys doing better than the later ones (see Figure 3).
In Grade 8 there was only one significant interaction; the one between sex,
pubertal stage and SES-group (p < 0.05). Regarding pubertal stages, late maturing
boys (except for the prepubertal boys) were performing better than the early
maturing ones, while there were no differences between early and late maturing girls.
Among the girls there were no SES-group differences, while among the boys SES-
group II was doing better than SES-group I.
On the test measuring level of cognitive thinking - "Similarities"- there were no
significant differences or interactions between sex, pubertal stage and SES-group in
Grade 4.
In Grade 6, however, there were significant differences between children of
different pubertal stages - the early maturing ones performed better. At the same
time, there was a significant interaction between sex and pubertal stage (p < 0.05),
which is illustrated in Figure 3.
In Grade 8, there were no significant differences on the test "Similarities"; either
between sexes, the pubertal stages or SES-groups. Two significant interactions
were, however, found. One between sex and pubertal stage (p < 0.05), which is
illustrated in Figure 3; the other interaction between pubertal stage and SES-group
(p < 0.01) indicating that later maturing children in SES-groups I and II performed
better than the earlier maturing ones, while there were no differences between early
and late maturers in SES-group III.
4. DISCUSSION
The results of this cross-sectional study of Swedish schoolchildren lend credence to
the hypothesis that children who are early maturers in the physical sense are prone
to be early maturers in at least some aspects of their mental development also.
A previous study, at that time longitudinal, of Swedish urban schoolchildren
born in 1955 also relating mental ability and school achievement to early and late
maturation (Westin-Lindgren, 1982) gave results quite consonant with those of the
present investigation. However, there were also - apart from the difference in the
design of the two studies - some other disparities between them.
For one thing, the criterion for defining late and early physical maturation was in
the longitudinal 1955-study age at maximum height velocity during puberty (the so-
called Peak Height Velocity age or PHV-age), while in the cross-sectional 1980-
study the criterion was ratings of pubertal stages in five stages (from prepubertal to
adult stage). This means that in the former study, the criterion was the same for
boys and girls and the reliability of the measurement regarding maturational rate was
higher. The relationship between these different criteria of physical maturity is,
188 CHAPTER 16
however, reasonably high (cf. e.g. Lindgren, 1978) indicating that, if a child is early
according to one criterion, he or she is probably also early according to the other
criteria. For boys the highest correlation coefficient is r = 0.79 between age at
pubertal stage IV and PHV-age. For girls the highest correlation is r = 0.63 between
pubertal stage III and PHV-age.
Also the mental ability tests used in both studies were administered in different
Grades. In the longitudinal study they were administered only in Grade 5 (11 - 12
years), while in the cross-sectional study these tests were administered in Grade 4
(10 - 11 years), Grade 6 (12 - 13 years) and Grade 8 (14 - 15 years). In addition a
new type of test measuring cognitive levels of thinking according to Piaget's theory
was administered in all three Grades in the cross-sectional study.
In spite of the differences in design the main results from both studies, however,
give a similar pattern. Verbal ability as measured by the test "Opposites" did not
differ between early and late maturers in Grade 4 (age 10 - 11 years) according to the
present study. In Grade 5 (11 - 12 years), however, according to the former
longitudinal study early maturing girls were performing better than late maturing
girls on this test. In Grade 6 (12 - 13 years) there were no differences between early
and late maturers; neither in Grade 8 (14 - 15 years) according to the present study.
Girls in Grade 8, however, performed better than boys.
Regarding logical-inductive ability as measured by the test "Letter groups" early
maturers were performing better than late maturers (especially the girls) in Grade 4
according to the present study. In addition SES-group I was performing better than
SES-groups II and III. However, according to the former longitudinal study, there
were in Grade 5 no differences between early and late maturers or between SES-
groups. In Grade 6 again, according to the present study, early maturers (and this
time especially the boys) were performing better than late maturers. In Grade 8 there
were no longer any differences between either pubertal stages or between SES-
groups per se.
The cognitive test "Similarities" did not reveal any differences between either
sex, pubertal stages or SES-groups in Grade 4. There was, however, in Grade 6 a
rather clear relationship between pubertal stage and test results for both boys and
girls showing continuously better results from prepubertal stage I to adult stage V.
There was though an interaction between sex and pubertal stage indicating a "dip" in
the trends at pubertal stage III for boys and pubertal stage IV for girls followed by a
catch-up at pubertal stage IV and V respectively. This interaction makes one wonder
whether the timing of PHV-age is in some way related to it. Boys were performing
lower than girls at pubertal stages I, II and III, but after pubertal stage III surpass the
girls at pubertal stages IV and V, that is to say at a time when boys have their
PHV -age. Results from the earlier longitudinal study showed that boys on a mental
arithmetic test scored lower than girls in Grades 5 and 6, but then caught up and
surpassed the girls in Grade 7 (13 - 14 years) when in general boys have their PHV-
age (Lindgren, 1979). In Grade 8 there were in the present study no significant
differences between pubertal stages regarding test scores on the test "Similarities".
There were, however, some interactions between; on one hand sex and pubertal stage
- on the other pubertal stage and SES-group, which were difficult to give
meaningful interpretations.
The results from the present study added to the results from the former
longitudinal study strengthen the impression that tempo of physical maturation
MENTAL ABILITY AND COGNITIVE THINKING 189
during puberty plays an important role when explaining the success of children's
performance at school, most evident from lO to 14 years of age. After the age of 14
years the direct effect of physical maturation seems to have a less important role - at
least as concerns mental ability and cognitive thinking - and the effects of sex and
socio-economic background take over. However, long-term effects might be
expected, if the late maturing children's self-esteem might have been negatively
affected during the pubertal period. A more detailed attempt to interpret the various
results within their social context as well as the educational implications have been
given by Lindgren (1988, 1995).
5. REFERENCES
Anastasi, A., 1958, Differential Psychology. Individual and Group Differences. 3rd ed. (New York:
Mac Millan).
Boas, F., 1941, The relation between physical and mental development. Science, 93, 339-342.
Douglas, J.V.B., and Ross, J.M., 1964, Age at puberty related to educational ability, attainment and
school leaving age. Journal of Child Psychology and Psychiatry, 5, 185-196.
Emanuelsson, I., and Svensson, A., 1986, Does the level of Intelligence Decrease? A Comparison
between Thirteen-Year Olds Tested in 1961, 1966 and 1980. Scandinavian Journal of Educational
Research, 30, 25-37.
Epstein, H.T., 1974, Phrenoblysis. Special brain and mind growth periods. Developmental
Psychobiology, 7, 207-224.
Haernqvist, K., 1960, Individuella differenser och skoldifferenser. [Individual Differences and School
Differences] (Stockholm: SOU 1960: 13).
Husen, T., 1974, Talent, Equality and Meritocracy. Availability and Utilization of Talent. Plan Europe
2000. Project I: Educating Man for the 21st Century. Vol. 9. (The Hague: Martinus Nijhoff).
Inhelder, B., and Piaget, J., 1958, The Growth of Logical Thinking from Childhood to Adolescence
(London: Kegan Paul).
Johannesson, I., 1?74, How does puberty}nfluence school achievement. In Compte-rendu de la XIIe
Reunion des Equipes Chargees des Etudes sur la Croissance et Ie Deve10ppement de I'Enfant
Normal (Paris: Centre International de l'enfance) p. 243.
Kelly, A., 1978, Girls and Science. lEA Monograph Studies No.9. (Stockholm: Almqvist & Wiksell
International).
Kohen-Raz, R., 1977, Psychobiological Aspects of Cognitive Growth (New York: Academic Press).
Lindgren, G, 1976, Height, weight and menarche in Swedish urban schoolchildren in relation to socio-
economic and regional factors. Annals of Human Biology, 3, 501-528.
Lindgren, G., 1978, Growth of schoolchildren with early, average and late ages of peak height
velocity. Annals of Human Biology, 5, 253-267.
Lindgren, G., 1979, Peak velocities in height and mental performance. A longitudinal study of
schoolchildren aged 10-14 years. Annals of Human Biology, 6, 559-584.
Lindgren, G., 1988, Psycho-Social Aspects of Growth with Special regard to the relation between
PhysicaVPhysiological and MentaVCognitive Growth. Collegium Antropologicum, 1,47-66.
Lindgren, G., 1995, Socio-economic background, growth, educational outcome and health. In Essays
on Auxology, edited by R. Hauspie, G. Lindgren and F. Falkner. (Welwyn Garden City:
Castlemead Publications) p. 408.
Ljung, B.-O., 1965, The Adolescent Spurt in Mental Growth (Stockholm: Almqvist & Wiksell).
Maccoby, E.E., and Jacklin, C.N., 1974, The Psychology of Sex Differences (Stanford, California:
Stanford University Press).
Mitchell, J.R., 1980, Male Adolescents' Concern about a Physical Examination conducted by a Female.
Nursing Research, 29, 165-169.
Muuss, R.E.H., 1970, Adolescent development and the secular trend. Adolescence, 5, 267-284.
Nisbet, J.D., and Illsley, R., 1963, The influence of early puberty on test performance at the age of
eleven. British Journal of Educational Psychology, 33, 169-176.
Sandgren, B., 1974, Kreativ utveckling. En empirisk studie av kognitiv utveckling samt en kritisk analys
av intelligensbegreppet [Creative development. An empirical study of cognitive development and a
critical analysis of the concept intelligence] (Stockholm: Almqvist & Wiksell).
190 CHAPTER 16
SAS User's Guide: Statistics., 1985, Version S Edition, Chapter 20: The GLM Procedure. (Cary, N.C.:
SAS Institute Inc) p. 433.
Shuttleworth K.F., 1939, The physical and mental growth of girls and boys age six to nineteen in
relation to age at maximum growth. Monographs of the Society for Research in Child
Development, 4, No.3.
Tanner, 1.M., 1961, Education and Physical Growth. (London: University of London Press).
Tanner, 1.M., 1962, Growth at Adolescence. 2nd ed. (Oxford: Blackwell Scientific Publications).
Tanner, 1.M., 1989, Foetus into Man. 2nd ed. (Ware:: Castlemead Publications).
Tanner, 1.M., 1999, Retrospective: the growth and development of the Annals of Human Biology: a 2S-
year retrospective. Annals of Human Biology, 26, 3-18.
Westin-Lindgren, G., 1982, Achievement and mental ability of physically late and early maturing
schoolchildren related to their social background. 10urnal of Child Psychology and Psychiatry, 23,
407-420.
Westin-Lindgren, G., 1984, Physical and mental growth controlling for social background. In Human
Growth and Development edited by 1. Borms, R. Hauspie, A. Sand, C. Susanne, and M. Hebbelinck
(New York: Plenum Publishing Corporation) p. 70S.
MENTAL ABILITY AND COGNITIVE THINKING 191
o
If you divide a circle like this ...
CD C£ CJD
... you can get halves
that are exactly alike.
o o--~©
If you divide an apple like this ...
... you will see that the halves
aren't exactly alike. But you can
still say that they are alike
O 0,
Compare an apple and a pear!
You can see that they are rather unlike L3 ttp'I-~\
each other. A
But they are also alike in many ways.
Many things are like that. For example, take a mushroom
and an umbrella. How do you think they are alike?
~ / , / A sled and a pair of skies can also be alike in many ways.
6tl/ / / What similarities can you find between them? Write some
down here: .. ........ .. ....... ........ ....... ..... ... ... ....... ........ ...... . .
You understand now that things can be alike in many ways. You can't say that ONE
such similarity is the only right one! There are many similarities, and ALL are
correct, each one in its own way.
Now, if you have several similarities and are going to choose between them, you
usually think that one is better than the others. When people choose between
similarities, they usually choose very different ones. But none can be wrong, since
all of the similarities are right.
Now, please tell me what YOU think is the best similarity between a bun Q and a
loaf of bread Ch
In the squares below, we have written down some similarities, which you can
choose from:
Best Worst
1 2 3 4 5 similarity similarity
can be are baked are light contain taste good
eaten in an oven brown yeast with butter
-- -
As you can see, we have numbered the similarities from 1 to 5. Now, choose first
the similarity you think is the BEST one! Write the figure above that similarity in
the square "Best similarity".
Then, choose the similarity you think is the WORST one! Write this figure above
that similarity in the square "Worst similarity".
On the next page, you will find a lot of items that are like this one. Choose the best
and the worst similarities for each item. Don't worry about what your classmate
chooses. Write only what YOU YOURSELF think is the best and the worst similarity.
192 CHAPTER 16
APPENDIX
Write your name, class and school here before you start to fill out this answering
sheet!
Best Worst
1 2 3 4 5 similarity similarity
APPENDIX
Best Worst
1 2 3 4 5 similarity similarity
An arm- have four have the you can are furni- are com-
chair legs same sit in them ture fortable
A sofa shape
A one are money you can are reaso- you have have pic-
dollar bill pay for nable them in tures of
A five things weekly your famous
dollar bill with them allowance wallet men on
them
A cap are round you can protect are head- are easy
A hat have them you gear to forget
on your
head
THE END
194 CHAPTER 16
APPENDIX
10 I
Grade 4: Girls
8 ...........
•
ti D Pub. stage I
,..0 Pub. stage II
~
6 ··········· Pub. stage III
~
d)
on
ro 4
ti
:>
<t:
2
2 3 4 5
Cognitive level
Grade 4: Boys
•
d)
.0
E
a 61-···· ····1 Pub. stage II
~
~ 4
d)
:>
<t:
2
0
0 2 3 4 5
Cognitive level
APPENDIX
Grade 6: Girls
16 ...........
•
~ 0 Pub. stage I
..0 Pub. stage II
~ 12 ...........
~ Pub. stage III
0
~
8 Pub. stage IV
r:!
C1.)
8 ··········· D Pub. stage V
;:-
<
4
0
0 2 3 4 5
Cognitive level
2
Grade 6: Boys
I-<
C1.)
D Pub. stage I
..0 • Pub. stage II
§ 12 ~ Pub. stage III
o
B Pub. stage IV
&C1.)
o Pub. stage V
;:-
<
0' 0 2 3 4 5
Cognitive level
APPENDIX
Grade 8: Girls
~
12 ~···········
D Pub. stage I
1
d 8 • ..........
•
~
a
Pub.
Pub.
Pub.
stage II
stage III
stage IV .......................................................................... ~
& D Pub. stage V
0
0 2 3 4 5
Cognitive level
16
I
Grade 8: Boys
•
12 ...........
1-0
Q)
D Pub. stage I
.0 Pub. stage II
aE 8 ··········· ~
g
Pub. stage III
Pub. stage IV
&
Q)
;>
0 Pub. stage V
< 4
! !
0' 0 2 3 5
Cognitive level
APPENDIX
Table 1. Results concerning three-way ANOVA (GLM-procedure) regarding the verbal test
"Opposites" in relation to sex, pubertal stage and SES-groups in
Grade 4 (10-11 years):
N= 76
Table 2. Results concerning three-way ANOVA (GLM-procedure) regarding the verbal test
"Opposites" in relation to sex, pubertal stage and SES-groups in
Grade 6 (12-13 years):
N=81
APPENDIX
Table 3. Results concerning three-way ANOVA (GLM-procedure) regarding the verbal test
"Opposites" in relation to sex, pubertal stage and SES-groups in
Grade 8 (14-15 years):
N=84
APPENDIX
APPENDIX
APPENDIX
1. INTRODUCTION
Children tend to resemble their parents in stature, body proportions, body
composition, and rate of development. It may be assumed that barring the action of
obvious environmental influences on growth (such as chronic illness or long-term
malnutrition) these resemblances reflect the influence of genes that parents contribute
to their biological offspring. A study published by Prokopec and Lhotska (1989),
based on a sample of 81 boys and 78 girls, is an example of this view. The subjects,
all from Prague, were measured annually from birth to age 20 years. The Preece-
Baines growth curve was fit to the longitudinal data of each subject. From these
fitted curves for all the boys and girls, the three tallest, the three shortest, the three
slowest maturing, and the three fastest maturing of each sex were selected. None of
these extreme cases was known to have any major chronic or acute diseases. Neither
the subject's history of common childhood diseases, nor the occupation of the fathers
had an effect, positive or negative, on growth and development. In contrast, the mid-
parent height did predict the adult stature of offspring. Mid-parent height is the
average of the stature of the mother and the father. Inspection of the Preece-Baines
curves showed that tall or short stature at age 20 could be predicted from stature at
age four years. The positive impact of mid-parent stature on offspring growth and the
predictability of adult height from stature at age four are prima/acia evidence for the
role of heredity. Moreover, these findings attest to the early establishment of
individual patterns of growth and their stability over time.
Other well-known research strategies are used to demonstrate the genetic
determination of human growth. These strategies include the study of monozygotic
and dizygotic twins, correlations in growth between biological relatives (non-twins),
and the effects of genetic abnormalities on growth (reviewed in Bogin 1999). Studies
such as these lend support to the concept of a "genetic potential" for body size, body
composition and body proportions. The term "genetic potential" usually means that
every human being has a genetically determined upper limit to adult stature, the ratio
205
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 205-221.
© 2001 Kluwer Academic Publishers.
206 CHAPTER 17
Figure 1. Comparison of European and African Olympic athlete physiques showing the
relatively longer legs of Africans. 1n each of the four comparisons the European athlete is
on the left. This figure was re-imagined from a photograph in The Cambridge
Encyclopedia of Human Growth and Development (Ulijaszek 1998, p. 360). Professor
Tanner supplied the original photograph).
208 CHAPTER 17
The photographs were originally published in Tanner (1964), and are of Olympic
athletes. Olympic athletes are a highly selected group of people. The body shape of
Olympic athletes, including the ratio of leg length to sitting height, is strongly
correlated with the type of sport performed. Moreover, Olympic athletes constitute
only a very small segment of humanity, far smaller in number than would be
required for reliable and accurate statistical analysis. Olympic athletes, then, are in no
way representative of the general population in terms of biology or behaviour. Yet, a
version of the original photographs has been reprinted in several textbooks,
including both editions of Eveleth and Tanner's Worldwide Variation in Human
Growth (1976, 1990), and is used to demonstrate genetic differences in growth
potential between human populations.
The concept of "genetic potentials" in growth and body shape of human
populations is well entrenched in the field of human growth research. Some scholars
take the concept of "genetic potentials" as a given and use it without definition or
justification from research, that is, without reference to the scientific literature.
Interested readers may peruse recent issues of journals dealing with human biology,
physical anthropology, paediatric medicine, and related fields, where they will find
the concept of "genetic potential" used in this way.
The problem with this casual usage is that the assumption upon which it is
based is not true. Existing research shows that adult stature, body proportions, and
body composition are highly plastic. One of the clearest examples of the plasticity
of human body proportions comes from Tanner et at. 's (1982) studies of relative lag
length of Japanese children. In 1957 the relative leg length of adult Japanese was
significantly less than that of Northern Europeans, and it was assumed that this
difference was due mostly to genes. By 1977 the two ethnic groups had become
virtually identical in shape. This invalidated the genetic explanation, which was
replaced by an explanation favouring improvements in the physical and social
environment of Japan. Research conducted in China (Zhang and Huang 1988),
Argentina (Bolzan et at. 1993), Poland (Wolanski 1979), and Mexico (Gurri and
Dickinson 1990, Wolanski et at. 1993, Wolanski 1995, Siniarska 1995) shows that
the development of body proportions during the growing years is exquisitely
sensitive to the quality of the local environment (reviewed in Bogin 1999 and
discussed below). In these studies boys and girls of identical ethnicity and genetic
background are compared, and children from lower SES families are significantly
shorter in leg length than children of higher SES families.
In this article we re-examine the existing literature to see to what extent ethnic
and population differences in body-proportions which are assumed to be of a genetic
nature are, in fact, better explained by environmental plasticity in growth. We also
present the findings of a new study of body proportion change in the children of
Guatemalan immigrants now living in the United States. Finally, we make use of
experimental findings on the determinants of fluctuating asymmetry in insects to
develop an ecological model of body-proportion development.
human growth studies. The authors present data for several measures of body
proportions. Here we re-analyse the data for relative sitting height, which is
calculated as [(sitting height/stature) x 100]. Relative sitting height is a ratio that
expresses the percent contribution of sitting height -- the length of the head, neck
and trunk -- to total stature (Lohman, Roche, and Martorell 1988). In practice, this
ratio is most often used as an indication of leg length differences between individuals
or populations. The reason for this is due to the cephalo-caudal gradient of growth,
which means that during the years of growth the head and trunk are always closer to
their final adult size relative to the legs. The sitting height ratio changes with age as
the legs begin to grow relatively faster than the head and trunk of the body.
From the data found in Eveleth and Tanner (1976, 1990) we analysed a total of
874 samples of relative sitting height of boys and girls, measured at different ages
and representing populations from many countries. The samples include people
living on most of the major continents and islands of the world. When discussing
the relative sitting height data, Eveleth and Tanner focus on only a few of these
samples, which they consider to be representative of four major geographic regions
of the world. The 1976 edition of their book presents samples from London
(Europe), from Ibadan, Nigeria (Africa), Hong Kong (Asia), and of Australian
Aborigine origin (Australia). In proportion to sitting height, the Australians had the
longest legs followed, in order by Africans, Europeans, and Asians. Expressed
quantitatively, "at a sitting height of 60 cm, for example, London boys have leg
lengths averaging 43 cm, Ibadan boys 53 cm and Australian Aborigine boys 61 cm"
(Eveleth and Tanner 1976 p. 229). In the 1990 edition of their book they analyse
data from Bergen (Europe), the People's Republic of China (Asia), "Afro-Americans"
(Africa) measured for the NHANES II study (a national study of health and nutrition
in the United States), and the Australian Aborigines from 1976 (Australia). The
1990 analysis finds almost the same differences in leg length as found in 1976.
By organising the samples in this way, Eveleth and Tanner are employing a
geographic "racial" typology (Garn and Coon 1955, Gam 1971). Garn (1971, p. 17-
18) states that, "To a large extent the geographical races of mankind coincide with
the major continents and ... may also be spread over major island chains, as is evident
in the Pacific today." Over the past century, many researchers have assumed that the
body proportion differences between geographic populations are explainable only in
terms of a genetic or "racial" model. The population difference in body proportions is
usually considered to be an adaptation to regional climates, for example, long arms
and legs in hot and humid regions and short appendages in cold regions. There is
considerable evidence to support the correlation between climate and body
proportions (Roberts 1953 is the classic study). However, no genes for the
determination of body proportions are known and a plausible genetic mechanism that
might account for population differences in body shape has never been formulated.
We use quote marks when writing the word "race" because that term implies a
biologically definable group of people. Definable biological distinctions between so-
called "races" do not exist at the genotypic or phenotypic level (Bogin 1993, Lasker
1999). "Race" does have some value as a shorthand term to categorise people into
groups that differ in economic opportunities, social organisation and resources, and
political power. In the United States those people who are classified as "white"
enjoy, on average, greater socio-economic opportunities and political power than
those classified as "black." These differences have an impact on social resources that
210 CHAPTER 17
often translate into biological effects, such as patterns of growth and development
(see Bogin 1999 for a more general and comprehensive discussion of "race" and
human growth).
The analysis presented by Eveleth and Tanner compares four samples from
different geographic areas. These areas conform to the four major world regions
defined by many of those researchers using a "racial" typology to divide humanity --
Africa, Asia, Australian, Europe. Eveleth and Tanner (1990, p. 188) attribute the
differences in relative sitting height and other body proportions to "racial
differences." Indeed, all of the growth data presented in both editions of Worldwide
Variation in Human Growth are organised into categories based upon geographical
"races." These "races" are Europeans, Africans, Asian, Australian Aborigines and
Pacific Islanders (including New Guineans and Maoris of New Zealand), Indo-
Mediterraneans (includes Algeria, Egypt, Ethiopia, India, Iran, Israeli Kurds and
Yemenites, Pakistan, Saudi Arabia, Turkey, and Yugoslavian gypsies), and a final
group of "inter-racial crosses", such as Cuban mulattos, South African Coloureds,
and Mexican Mestizos. The geographic "racial" classification is reinforced in their
work in that Eveleth and Tanner are more concerned with the putative geographic
origin of the samples than with the location at which they were measured. For
example, Black children measured in Tanzania and Black children measured in
Washington, DC are considered to be "African race." Native Canadians (Indians of
Canada), the Maya of Guatemala, and Chilean Indians are considered to be "Asian
race." Any samples of children in the United States with light skin colour and
claiming European origin are considered to be of the "European race."
Table 1. Variables used to re-analyse Eveleth and Tanner's data, the scores assigned to
each variable and the meaning of these scores.
Variable Score
AGE 1 to 18 - indicates the chronological age of a sample; e.g. a
score of 7 is for individuals between 7.00 and 7.99 years old. A
score of 20 indicates age 20 years old or older.
SEX 1 = male and 2 =female
RACE 1 = Origin in Australia, New Zealand, or Papua New Guinea, 2
= Origin in Africa, 3 = Origin in Europe, 4 = Origin in Asia,
includes Native Americans (North, Central, and South
American Natives).
REGION I = Europe, 2 = North America (Canada and United States), 3 =
Central America (Mexico to Panama), 4 = South America, 5 =
Africa, 6 = Asia, 7 = Indo-Mediterranean, 8 = Australia, New
Zealand, Papua New Guinea, and Pacific Islands
WORLD 1 = Industrialised/developed nations, 2 = former Soviet Union
and its satellites, 3 = non-industrialisedllesser developed
nations, 1.5 = Hong Kong, Singapore, Southern Europe
SES 1 = very high SES, 2 = high SES, 3 = middle SES in the
industrialised nations, 4 = urban poor/working poor, 5 = rural
poor
SES x WORLD Multiplication product of SES and WORLD
The WORLD variable indicates that a sample was measured in a first world,
second world, or third world country. First world countries in the Eveleth and Tanner
database are those of North-western Europe, the United States, Canada, Japan,
Australia, and New Zealand. Second world nations in the database are those that
belonged to the former Soviet Union and its satellite republics. Third world nations
in the database are those of Africa, Asia (except Japan), Latin America, the Pacific
Islands, New Guinea, and India. Some nations in Eveleth and Tanner's database were
most difficult to categorise. We assigned the value of 1.5 to Hong Kong, Singapore,
and southern Europe (Spain, Italy, and Greece) indicating that these nations were
halfway between first and second world in development indicators at the time the
growth data were collected. The WORLD variable is intended to assess the effect of
the general level of industrial development and standard of living on body
proportions.
The SES variable is an estimate of socio-economic status for each sample
varying from a high of 1 to a low of 5. Scores of 1 or 2 were assigned to samples
living in any country if the sample was described as "very high" or "high" SES. A
score of 3 is for middle SES samples in the industrialised nations of the first and
second world. A score of 4 is for urban poor and a score of 5 is for rural poor.
Parental occupation and parental education, especially of the father of a child, are
usually used to estimate SES. But, such information is not always known, and this
212 CHAPTER 17
makes the SES variable the most difficult to assign with accuracy. We tried to use
the SES information provided in each of the original publications cited by Eveleth
and Tanner. Most often, the authors of these papers describe their samples as "low
SES," "middle SES," or "high SES." However, not all of the original papers
included measures of SES for their sample. Also, we could not locate all of the
original papers. Our assignment of a SES score to each sample is, therefore, a
mixture of quantifiable information and qualitative assessment based on our best
estimate of the general socio-economic status of the sample.
The SES x WORLD variable is the multiplication product of the variables SES
and WORLD. We included this variable based on existing empirical research
showing an interaction between SES and WORLD on human growth (e.g.
Henneberg and Van Den Berg 1990, Bogin 1999). For example, low SES Maya
children growing up in the United States, a first world nation, are significantly taller
and heavier than low SES Maya children of the same ages growing up in their third
world homeland of Guatemala (Bogin and Loucky 1997).
For analysis, each of these variables was entered into a ridge regression model.
Ridge regression is used when multicollinearity exists between the independent
variables. A test for multicollinearity found that the variables WORLD and SES x
WORLD were highly correlated with the other variables. Ridge regression reduces
the effects of such high correlations by adjusting the regression model for more
reliable beta coefficients.
2.2 Results
Table 2. Summary of the stepwise ridge regression model for relative sitting height
(lambda set at 0.10). R = 0.684, R2 = 0.465, F(4, 869) = 191.03, p < 0.000, Standard
error of estimate = 2.083
2.3 Discussion
Eveleth and Tanner state that, "The ultimate size and shape that a child attains as an
adult is the result of a continuous interaction between genetical and environmental
GENETICS OF BODY PROPORTIONS 213
influences during the whole period of growth" (1990, p.176). We agree with this
statement. Our re-analysis of Eveleth and Tanner's data helps to refine the role of
"genetical and environmental influences." Our regression analysis shows that
geographic RACE does contribute to variation in body proportions, but at a
relatively low level. Explaining only 3.6% of the total variation, RACE contributes
much less to the variability in body proportion than is assumed by many researchers
and the general public. The socio-economic variable SES, also contributes a small
and significant amount to the variance in body proportion. Indeed the relative
contribution of RACE and SES to the variance in body proportion are statistically
equal in our analysis, as a test for equality of slopes of their regression coefficients
(the unstandardised betas) shows no differences (t = 0.94). It is important to note
that, due to the imprecise definition of both RACE and SES in our re-analysis, our
results must be taken with caution.
To better understand variation in body proportions it is necessary to have more
accurate control of genetic and environmental variables. Some studies of body
proportion have such control. For example, Ramos-Rodriguez (1981) shows that
Mexico City children of middle SES have relatively longer legs than low SES
Mexican children from Oaxaca. She also shows that genetics cannot account for this
difference. Genetics played no role in the secular trend of Japanese body proportions
studied by Tanner et al. (1982). Prior to 1960, the Japanese were considered to be
both a short stature and short-legged "race." Between 1960 and 1977 the Japanese, on
average, gained 10 cm in stature and almost all of this increase was in leg length.
After 1977, the average height of Japanese men and women continued to increase,
but at a slower rate (Takaishi 1995), and both leg length and sitting height seem to
have increased at about the same rate, at least for young women (Rojo et al. 1981).
Since 1990 there is little evidence for further increase in stature. Today, Japanese
have, on average, virtually the same body proportions as many European
populations.
Several researchers working in Argentina, Poland and Mexico report similar
findings on the plasticity of body proportions. A team of researchers working in
small towns located in the Province of Buenos Aires measured 569 boys and girls,
seven to 13 years old, attending several schools (Bolzan et al. 1993). The sample
was divided into groups according to age, sex, and occupational status of the father.
Both boys and girls with fathers of lower occupational status (lower SES) were
shorter, and especially shorter in leg length, than subjects of higher family SES. In
Poland (Wolanski 1979), improvements in living conditions in towns and villages,
such as nutrition and health care, are associated with increases in leg length relative
to stature. Similar results are reported from a series of Mexican studies that find that
the body proportions of boys and girls vary according to relatively small differences
in family SES. The authors of these studies (Wolanski et al. 1993, Wolanski 1995,
Siniarska 1995) measured children living in the Yucatan region, including both
ethnically Maya and non-Maya populations. All of the families were of, generally,
low SES, but children from families of slightly better economic means were longer-
legged than children from lower SES families. Other studies from Mexico find even
more subtle influences of life style on body proportions. The leg-Iength-to-stature
proportions of women living in Chiapas Mexico, all of low SES, differ according to
socio-economic status, ecological region, and demography (Gurri and Dickinson
1990). All of the 421 women studied were 20 years old or older and all were of
214 CHAPTER 17
generally low SES. The sample was divided into four SES regions: 1) a region of
intensive export agriculture, 2) a region of cattle for meat production, 3) a region of
mixed agriculture and dairy herding, both for national consumption, and 4) a region
of subsistence agriculture. The authors also divided the sample into four ecological
regions: 1) Pacific coastal plain, 2) Sierra Madre mountains, 3) Central Valleys, and
4) Central Plateau. Finally a demographic division was made between those women
living in rural or urban areas. The authors found that 80% of the variance in stature
in the sample was due to SES region and that 20% of the variance was due to
ecological region. Almost all of the differences in stature within the sample were due
to variation in growth of the leg. Women from the SES regions of export agriculture
and the cattle raising area were the tallest. Women from the highland ecological
regions were, generally, the shortest. Similar findings are reported for children and
adults living in the Yucatan (Murguia et at. 1990, Dickinson et at. 1990). These
Mexican studies indicate that even within a generally low SES population, life style
differences exist and exert influence on body proportions.
3.2 Results
Mean values for height and sitting height ratio for each sample are shown in Figures
2 and 3. Anthropometric reference data from the NHANES I and II surveys of the
United States (Frisancho 1990) are used as a baseline for comparison in each graph.
Analysis of variance (ANOV A) was used to evaluate differences between samples.
GENETICS OF BODY PROPORTIONS 215
160
- Maya-Guat
150 • Maya-USA1992
- Maya-USA2000
E 140 - - - - - NHANES
E
.1::"
01)
130
.Qj
:r::
120
110
100
5 6 7 8 9 10 11 12 13
Age, years
Figure 2. Mean height of Maya children living in the United States measured in 1992
(Maya-USA 1992) and in 1999-2000 (Maya-USA 2000) compared with Maya children
living in Guatemala measured in 1998 (Maya-Guat) and the United States reference data
from NHANES 1 and 1/.
58
- Maya-Guat
57 - Maya-USA2000
- - - - - NHANES
0
.~
56
...
.1::01) 55
.Qj
..c:
01)
:::
54
·il
iZi 53
52
---
51
5 6 7 8 9 10 11 12 13
Age, years
Figure 3. Mean sitting height ratio of Maya children living in the United States measured
in 1999-2000 (Maya-USA 2000) compared with Maya children living in Guatemala
measured in 1998 (Maya-Guat) and the United States reference data NHANES 1 and 1/.
216 CHAPTER 17
After adjusting for the effect of age and sex, our results show (Figure 2) that the
Maya children of the 1999 and 2000 samples (abbreviated as "Maya-USA 2000") are
significantly taller (Figure 1) than the Maya samples measured in 1992 (abbreviated
as "Maya-USA 1992") and Maya children living in Guatemala (abbreviated as
"Maya-Guat"). The Maya living in the USA also have smaller sitting height ratios
than Maya living in Guatemala (Figure 3). All differences are significant at p < .01.
A smaller sitting height ratio generally indicates a child with relatively longer legs.
Compared with the NHANES references, all of Maya are shorter and have higher
sitting height ratios, i.e, relatively shorter legs.
3.3 Discussion
Our findings add further support to the literature on developmental plasticity in body
proportions. The results indicate that between 1992 and the present, there is a clear,
and positive, trend in growth of Maya children living in United States. The reasons
for this trend are likely due to improvements in the environment for growth. All
Maya in the USA have access to clean drinking water, health care, and education.
The Maya children living in Indiantown, Florida participate in school breakfast and
lunch programs. Nearly all of the Maya children have at least one parent with a
wage-earning job. All of this provides the Maya in the United States with a higher
standard of living than that found in Guatemala, as clean water, health care, education
and wage paying employment often do not exist for Maya living in rural Guatemala.
These health, economic, social, and nutritional changes are known to result in
greater stature. Our findings support the hypothesis that the increase in stature is due
mostly to relatively longer legs. We predict that the positive trend in height and
sitting height ratio seen between 1992 and 2000 will continue for the Maya in the
United States. Eventually heights and body proportions should achieve values that
approximate those for long-term residents of the United States.
4. CONCLUSION
In this article we have taken a critical look at the concept of genetic potential or
genetic determination of human body proportions. We find that population
differences in body proportions are influenced very little by genetic background, at
least as imputed from geographic, ethnic, or "racial" categorisations. Our findings
stand in sharp contrast to the opinion found in several of the most widely consulted
books on human growth. In a more general sense, the whole concept of "genetic
determination" in human growth is seriously flawed. That notion implies that the
flow of information about how any human trait is developed, be it height, body
fatness, personality, or intelligence, originates in the DNA and then unfolds into the
phenotype. Within this scenario, one may allow for a greater or lesser amount of
environmental influence on the phenotype, but the flow of information basically
begins with the DNA and moves one way.
The roles of DNA in human development are much more complex, and often
much less direct, than this. Genes do not directly cause growth and development.
Rather, the many proteins that genes produce, which are mediated by the endocrine
and neurological systems, regulate the expression of a genetically inherited pattern of
GENETICS OF BODY PROPORTIONS 217
growth. The physical and social environment also mediates growth. In the case of
the Prague study discussed in the introduction of this article, it is important to note
that the parents and children were living in the same households and, therefore,
shared a very similar environment. It is too simplistic, therefore, to ascribe the
similarities in growth between parents and offspring to genes alone.
The interactions between genes, hormones, and the environment may flow in all
directions. A marvellous example of the interaction of genes, proteins, and the
endocrine system may be seen in the action of homeobox genes and Hox genes. The
description and elucidation of homeobox and Hox genes is one of the most important
advances of molecular biology of the past two decades. The homeobox is sequence of
180 DNA base pairs that codes for a 60 amino acid segment of a protein. First
discovered in the genome of the fruit fly, Drosophila, homeobox sequences are found
in all eukaryote organisms so far examined. These highly conservative DNA
sequences -- the same homeobox sequences are found in organisms as diverse as
hydra, nematodes, all arthropods (the group that includes insects) and all chordates
(the group that includes human beings) -- produce proteins that regulate the
expression of other genes, " ... and control various aspects of morphogenesis and cell
differentiation" (Mark et al. 1997, p. 421). Hox genes are a category of homeobox
genes that encode transcription factors (Holland and Garcia-Fernandez 1996), which
are proteins that initiate and regulate the conversion of the DNA code to the RNA
sequence that is used to make amino acid polypeptide chains.
In multicellular animals, homeobox genes act to delimit the relative positions of
body regions, for example the head, thorax and abdomen of insects, or the general
body plan and limb morphology of vertebrates. Homeobox genes seem to have their
greatest impact during the earliest stages of development. The proteins that
homeobox genes produce are needed to regulate the expression of other DNA to,
" ... sculpt the morphology of animal body plans and body parts" (Carroll 1995, p.
479). The DNA affected by homeobox proteins will, in tum, produce other proteins
that mediate cellular differentiation, growth, and development. These "down stream"
proteins do not act alone. Some of them must combine via a process called
molecular zipping before they have any effect on a given segment of DNA
(McKnight 1991). These and other proteins need an appropriate environment to have
any effect. In placental mammals, the biochemical environment of the egg cell and, a
bit later in time, of the mother's womb and the placenta, provide a host of factors
needed for growth, including nutrients and hormones.
Throughout life, the endocrine system often provides the necessary biochemical
environment for gene action. The human adolescent growth spurt, for example,
requires adequate amounts of two hormones, growth hormone and testosterone (boys)
or oestradiol (girls), to be secreted into the blood stream. Without these two
hormones the genes that regulate growth of skeletal, muscle, and adipose tissue will
not increase enough in activity to produce the growth spurt. The endocrine system
also responds to the influence of many environmental factors that affect human
development. Under-supply or over-supply of many nutrients, such as vitamin A or
D and folate, can have major effects on the growth and development of tissues,
organs, and the body as whole. These nutrients influence growth via their effect on
the regulation of DNA expression, protein synthesis, and hormonal regulation (see
Bogin 1999 for several additional examples). Because it is situated between the
action of genes and the external environment, the endocrine system serves as a
218 CHAPTER 17
mechanism that unifies the genes we inherit and the environments in which we live
to shape the pattern of growth of every human being.
............. •
.." ....
4
• .. .'
'
~ '
gf
.~
3
.E
~
&.
~
.~
2
~
o
1+1--~--r-~--~--r-~--~--~~--4
25 35 45 55 65 75
dry weight of body (mg)
Figure 4. Regression of dry weight per forewing (the average of left and right sides) on dry
weight of the body for butterflies from which one, both, or no hindwing imaginal discs
had been removed (from Klingenberg and Nijhout 1998, plJ 36, with permission of the
author)
GENETICS OF BODY PROPORTIONS 219
5. REFERENCES
Bogin, B., 1993, Biocultural studies of ethnic groups. In Research Strategies in Human Biology, edited
by G. W. Lasker and CGN Mascie-Taylor (Cambridge: Cambridge University Press) pp. 33-61.
Bogin, B., 1999, Patterns of Human Growth, 2nd edition (Cambridge: Cambridge University Press).
Bogin, B., and Loucky, J., 1997, Plasticity, political economy, and physical growth status of Guatemala
Maya children living in the United States. American Journal of Physical Anthropology, 102, 17-32.
Bolzan, A. G., Guimarey, L. M., and Pucciarelli, H. M., 1993, Crecimiento y dismorfismo sexual de
escolares segun la ocupacion laboral paterna. Archivos Latinoamericanos de Nutricion, 43, 132-
38.
Carroll, S. B., 1995, Homeotic genes and the evolution of arthropods and chordates. Nature, 376, 479-
485.
Dickinson, F., Cervera, M., Murguia, R., and Uc, L., 1990, Growth, nutritioual status and environmental
change in Yucatan, Mexico. Studies in Human Ecology, 9, 135-149.
Eveleth, P. B., and Tanner, J. M., 1976, Worldwide Variation in Human Growth. (Cambridge:
Cambridge University Press).
Eveleth, P. B., and Tanner, J. M., 1990, Worldwide Variation in Human Growth, 2nd edition,
(Cambridge: Cambridge University Press).
Frisancho, A. R., 1990, Anthropometric Standards for the Assessment of Growth and Nutritional Status
(Ann Arbor: University of Michigan Press).
Gam, S. M., 1971,Human Races, 3rd edition (Springfield, Illinois: Charles C Thomas Publisher).
Gam, S. M., and Coon, C. S., 1955, On the number of races of Mankind. American Anthropologist, 57,
996-1001.
Gibson, M. C., and Schubiger, G., 2000, Peripodial cells regulate proliferation and patterning
Drosophila imaginal discs. Cell, 103, 343-350.
Gurri, F. D. and Dickinson F., 1990, Effects of socioeconomic, ecological, and demographic conditions
on the development of the extremities and the trunk: A case study with adult females from Chiapas.
Journal of Human Ecology, 1, 125-138.
Henneberg, M. and Van Den Berg, E. R., 1990, Test of socioeconomic causation of secular trend:
stature changes among favored and oppressed South Africans are parallel. American Journal of
Physical Anthropology, 83,459-465.
Hojo, T., Takemoto, R., and Shinoda, K., 1981, The secular unchangeability in relative sitting height of
female Kyushuites. Journal of the University of Occupational and Environmental Health, Japan, 3,
203-5
Holland, P. W. H. and Garcia-Fernandez, J., 1996, Hox genes and chordate evolution. Developmental
Biology, 173,382-95.
Klingenberg, C. P. and Nijhout, H. F., 1998, Competition among growing organs and developmental
control of morphological asymmetry. Proceedings of the Royal Society London, 265, 1135-1139.
Lasker, G. W., 1999, Race. Microsoft Encarta Encyclopedia 2000 (Redmond, Washington: Microsoft
Corporation).
Lohman, T. G., Roche, A. F., and Martorell, R., 1988, Anthropometric Standardization Reference
Manual (Champaign, Illinois: Human Kinetics Publishers).
Mark, M., Rijli, F. M., and Chambon, P., 1997, Homeobox genes in embryogenesis and pathogenesis.
Pediatric Research, 42, 421-29.
McKnight, S. L., 1991, Molecular zippers in gene regulation. Scientific American, 264, 54-64.
Murguia, R., Dickinson, F., Cervera, M., and Uc, L., 1990, Socio-economic activities, ecology and
somatic differences in Yucatan, Mexico. Studies in Human Ecology, 9, 111-134.
Norgan, N. G., 1998, Body-proportion differences. In: Cambridge Encyclopedia of Human Growth and
Development, edited by S. J. Ulijaszek, F.E. Johnston, and M. A. Preece (Cambridge University
Press), pp. 378-379.
Prokopec, M. and Lhotska, L., 1989, Growth analysis of marginal cases of normal variation. Anthrop.
Kozl., 32, 65-79.
Roberts, D. F., 1953, Body weight, race, and climate. American Journal of Physical Anthropology, 11,
533-558.
Ramos Rodriguez, R. M., 1981, EI significado del miembro superior: una hip6tesis a considerar. Bolotin
de Medicina de Hospital de Infant Mexicana, 38, 373-377.
Roche, J, and Boe!l, E. J., 1999, Development (biology), Microsoft Encarta Encyclopedia 2000
(Redmond, Washington: Microsoft Corporation).
Siniarska, A., 1995, Family environment and body build in adults of Yucatan (Mexico). American
Journal of Physical Anthropology, Supplement 20, 196.
Takaishi, M., 1995, Growth standards for Japanese children-an overview with special reference to
secular change in growth. In Essays on Auxology, edited by R. Hauspie, G. Lindgren, and F.
Falkner (Welwyn Garden City, England: Castiemead), pp. 302-311.
GENETICS OF BODY PROPORTIONS 221
Tanner, J. M., Hayashi, T., Preece, M. A., and Cameron, N., 1982, Increase in length of leg relative to
trunk in Japanese children and adults from 1957 to 1977: comparison with British and with Japanese
Americans. Annals of Human Biology, 9, 411-23.
Tanner, J. M., 1964, Physique of the Olympic Athlete (London: George Allen & Unwin).
Ulijaszek, S.J., 1998, Between-population differences in human growth. In The Cambridge
Encyclopedia of Human Growth and Development edited by SJ. Ulijaszek, F.E. Johnston, and
M.A. Preece (Cambridge: Cambridge University Press) p. 361.
Wilsey, B. J. and I. Saloniemi, 1999, Leaf fluctuating asymmetry in tree-line birches (Betula pubescens
ssp. tortuosa): genetic or environmentally influenced? Oikos, 87, 341-346.
Wolanski, N., 1979, Parent-offspring similarity in body size and proportions. Studies in Human Ecology,
3,7-26.
Wolanski, N., 1995, Household and family as environment for child growth (Cross cultural studies in
Poland, Japan, South Korea and Mexico). In Human ecology: Progress Through Integrative
Perspectives, edited by S. D. Wright, D. E. Meeker, and R Griffore (Bar Harbor, Maine: The
Society for Human Ecology), pp. 140-152.
Wolanski, N., Dickinson, F., and Siniarska, A., 1993, Biological traits and living conditions of Maya
Indian and non-Maya girls from Merida, Mexico. International Journal of Anthropology, 8, 233-
246.
Zhang, X. and Huang, Z., 1988 The second national growth and development survey of children in
China, 1985: children 0 to 7 years. Annals of Human Biology, 15,289-306.
CHAPTER 18
1. INTRODUCTION
The Sarsuna-Barisha Longitudinal Growth Study was conducted between 1952
and 1966 by the late Professor S.R. Das with the assistance of Mr. D.P. Mukherjee,
Mrs. Lalita Bose and Mrs. Asha Das, and with the financial support of the
Anthropological Survey of India. The study was named after the two Bengali
villages, near Calcutta, where all children were examined. More than 500 boys and
girls, belonging to some two hundred families, were examined between birth and
adulthood over varying lengths of time (sometimes up to 14 years). The study is
most remarkable by the fact that it was one of the first longitudinal growth surveys
in India, including 22 anthropometric measurements all taken by one single
measurer, Professor S.R. Das himself, using standardised anthropometric techniques.
Professor Das got thorough training and practice in anthropometry under the able
guidance of Dr. E.C. Biichi - then a visiting scientist in the Anthropological Survey
ofIndia.
In the early 70's, the hand-written data sheets were handed over to Professor I.M.
Tanner for further processing under his direct care and guidance at the Department of
Growth and Development of the Institute of Child Health in London. It was in late
1978 that these data was operationalised by one of the authors of the present chapter
(RCH) who had received a Postgraduate NATO Research Fellowship to work for one
year on this material under the guidance of Professor I.M. Tanner, Dr. N. Cameron
and Dr. M.A. Preece in the Institute of Child Health. Since then several results have
been published in co-authorship with Professor. S.R. Das. In 1985, Professor Das
set the raw data of the Sarsuna-Barisha study available to the scientific world through
a special volume published by the Anthropological Survey of India, Calcutta (Das
1985).
The aim of this contribution is to highlight the major outcomes of the analyses
of this most valuable longitudinal growth data.
223
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 223-236.
© 2001 Kluwer Academic Publishers.
224 CHAPTER 18
Number of sibs in 2 3 4 5 6 7 8 9
the family
Number 73 48 36 26 19 4 3 3 2
There were 303 boys and 260 girls, covering the age range 0.5 to 20 years. These
563 subjects belonged to 214 families; the frequency distribution of number of
sibships are shown in Table 1. Each child underwent a brief medical examination on
each measurement occasion and any with obvious illness or defect was omitted from
the record. Twenty-two anthropometric measurements were taken at six-monthly
intervals up to the age of five years, yearly from five to ten, six-monthly from lO to
14 and yearly thereafter. Only height will be presented in this paper. All children
were measured between 0700 and 0900 hours and all measurements, without
exception, were taken by S.R. Das throughout the entire period of the study.
Standard anthropometric instruments were used with the techniques described by
Martin (1928). For children below 2 - 3 years the measurement of stature was
substituted by the supine length from crown to heel, made on a special child-
measuring board. The true ages of the children were known in all cases as only
families having authentic birth records were included in the study; these records
included family horoscopes, notebooks and birth certificates. Target dates for
measurements were the birthdays or half-birthdays, and children were usually
measured within 15 days of the target. Over the whole material, attendance averaged
SARSUNA-BARISHA LONGITUDINAL GROWTH STUDY 225
six days late in boys and 7 days late in girls, and the standard deviation of the
difference between observational age and target age was about 14 days in both sexes.
Table 2 shows the number of children present in the survey for different lengths of
time. Table 3 shows the frequency distribution of the number of measurement
occasions per child.
Table 2. Number of children present in the study for different lengths of time (years)
3. METHODS
Mean height for boys between 2 and 20 years of age was estimated by fitting Preece
Baines model I directly to the raw height-for-age data (Preece and Baines 1978). The
model has the following mathematical formulation:
2(hi - he)
y = hI eso(t-e) + e Sl (t-e)
with five function parameters hI, he, so' SI, (), while y is height in em and t is
age in years. Although the model is originally designed to describe individual
growth, it is also very apt to fit population average height growth (Jolicoeur and
Pontier 1993, Zemel and Johnston 1994). The model was fitted by an efficient non-
linear least-squares technique developed by (Powell 1969) for the Harwell Subroutine
Library. This method is a compromise of three different algorithms for minimising a
sum of squares, namely Newton-Raphson, steepest descent and Marquardt (Hauspie et
al. 1980b).
Centile lines for attained height were estimated by applying the LMS-method
(Cole and Green 1992) to the residuals obtained after fitting the Preece Baines model
to the raw data, i.e. to the deviations of the height measurements from the fitted
mean curve. We used a FORTRAN computer program, kindly set available by T.
Cole and slightly adapted for the present analysis. The LMS-method is a powerful
and compact technique designed to construct growth standards, allowing for
departures from normality while at the same time allowing the centiles to be
calculated from the mean and standard deviation (Cole 1989). The program provides
values of L (the Box-Cox power transformation), M (the mean value), and S (the
226 CHAPTER 18
coefficient of variation) for distinct values of age. These values of L, M and Scan
then be used to calculate the desired centiles ( Pi) according to the formula:
Pi = M (1 + LSz)1L
4. RESULTS
100
P97
P90
170 P75
P50
P25
100 P10
P3
100
· .
......,, ... -.. ,,, ..........,,--- .... -,.---
: ' : :
.: .... -,---
.....·, ....... , ..... ---.---
:
,
E 13) ,
u l ' 1 1
E .:-------~-- .. ----~--
Ol
'(j)
I 12) -/,./.. ;/...;('......,.......;....... ;...... ) , ....... .j...... ;....... ) , ....... .j...... ,.
, .. ..
:, :, :: ::
... ~ ... ---j-...... +... --.. ~ ....... ~ ....... i-' -.... -t··· ····t·-· .. ··~
" "
, , " "
::,I;,1;lllil
110 ····'f----
100
.. ,,<..../-, ....... j ....... , ....... , ....... , ...... ; ....... , ....... , ....... j ........ ~ ....... j .....•. ; .......• , ....... j....... ;
; i : ; i l i j i : i i : i :
·····j·······,········;·······:·······j········t·······)······T······,·······)·······,
._... 1. .. /. .. :/. ..... ~........ ~ ....... )....... , ........ , ..
00 hI.;;./····;·······,········,··· ... ;....... , .......~ .......... ···1······ .... ·· ···· .... ········· ..,.... ···: ...... ··...... ··r .. ·.... ;.... ···1
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 2)
Age, years
170
100
150
140
oE 133
1:
Ol
·iii
111::11;;1;111]
1;+ ,····!ttl;:I.
I 12)
110
100
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 2)
Age, years
12
11
n; 10
Q)
~
E 9
0
E 8
·wCl
.r; 7
c
(/) 6
"E
Q)
E 5
~
0 4
.~
>.
3
~
Q)
>- 2
0
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 2)
Age, yea rs
n;
Q)
~o
E
Cl
w
:~ ::~ir'flii I
7 ,P50.
.r;
.~
(/) 6 1P1O :
"E ~.... j'"'<;.: .. ," ..... ~ ...~ "',' ...... ~ ... j / " ....... ~. '\.! .. \:" .. \ .. ~.\ .. !. --:- ... -.--~.- ..
Q)
5
E
~
o 4
.~
>.
3
~
Q)
>- 2
0
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 2)
Age, years
Figure 3. Centile distribution of yearly increments for height in boys (upper part) and
girls (lower part) from the Sarsuna-Barisha Longitudinal Growth Study.
230 CHAPTER 18
Table 4. Means and standard deviations (SD) for the five function parameters of Preece-
Baines modell and for a series of biological variables in 63 boys and 42 girls.
Boys Girls
mean SD mean SD
hI 165.72 5.18 151.77 4.68
he 153.40 5.18 141.90 4.79
so 0.099 0.023 0.130 0.023
sl 1.297 0.235 1.260 0.262
() 14.61 0.975 12.90 0.927
Ti: Age at take-off, years 10.47 1.47 9.32 1.11
HI: Height at take-off, em 129.70 6.08 121.31 6.40
Vi: Velocity at take-off, em/year 3.90 0.82 4.57 0.64
12: Age at peak velocity, years 14.26 1.00 12.35 0.99
H2: Height at peak velocity, em 150.59 5.04 138.19 5.04
V2: Peak height velocity, em/year 8.73 1.27 7.16 1.15
On the basis of this subsample of 63 boys and 42 girls, we could conclude that
these Bengali boys reach age at peak velocity 1.9 years later than girls but at ages
fairly similar to those seen in Western populations. A comparison of the Sarsuna-
Barisha data with longitudinal data from the Harpenden Longitudinal Study and the
London group of the International Children's Centre Longitudinal Study (Preece,
Tanner and Whitehouse, unpublished data), analysed in the same way it seemed that
the adult British boys were taller than Indian ones by 10 cm and that this difference
was almost entirely due to a difference in prepubertal growth, mean height at take-off
being 138 cm in the British boys and 129 cm in the Indian ones. The mean ages at
take-off in the two samples were practically identical and the total gain during
adolescence in the boys differed by less than a centimetre. The adult height of British
girls exceeded that of Indian girls by approximately 11 cm, of which 7 cm was due
to a difference in prepubertal growth and 4 cm to a greater adolescent gain in British
girls.
SARSUNA-BARISHA LONGITUDINAL GROWTH STUDY 231
Figure 4 shows the sex difference in height growth of Bengali adolescents on the
basis of mean-constant curves, i.e. the curves obtained by feeding the boys' and girls'
mean values of the 5 function parameters into the model (Hauspie 1989). Mean-
constant curves show the typical average growth pattern in the sample, and do not
suffer from the so-called phase-difference effect which smooths out the adolescent
growth spurt in cross-sectional growth data or in longitudinal growth data that is
treated cross-sectionally (unconditioned for tempo of growth). The velocity curve in
this graph is obtained by taking the mathematical first derivative of the distance
mean-constant curve.
170 Boys
100
150
E
u 140
1:
Ol
.(j)
1:Il
I
1~
110 10
9
100
8
00 7
6 <
CD
5"
5 Q.
4 ~
0
3 3
2
-<CD
~
According to Tanner et al. (1976) the sex dimorphism in adult stature of the
Indian boys can be decomposed into three additive parts:
• the difference in adolescent gain between boys and girls,
• the difference in pre-pubertal growth, measured as the size difference between boys
and girls at the girls' take-off,
232 CHAPTER 18
• the amount of growth achieved by the boys between the girls' and the boys' take-
off.
On a slightly different subsample, Hauspie et al. (1985) found that the sex
difference in adult height was greater in Indians than in British children (14.2 cm
compared to 12.0 cm). When expressed in percentage of the boys' adult height the
difference is even more pronounced in Indians than in their British counterparts. The
contribution of the difference in adolescent growth to the sex difference in adult
height was smaller in the Brits (2.0 cm) than in the Indians (5.2 cm). On the other
hand the contribution of the later onset of the pubertal growth spurt in boys was
bigger in Great Britain (7.9 cm) than in Indians (6.3). The sex difference in pre-
pubertal growth were fairly similar in British and in Indian children (respectively 2.1
and 2.7 cm). So it seemed that the greater sex difference in adult height among
Indians, compared to British children, was mainly due to a relatively greater
adolescent gain in Indian boys than in Indian girls.
Singletons Sibships
------------------------------
Number of children per family 1 2 3 4 5
Number offamilies 45 19 3 2 1
The results indicate that the timing and intensity of the adolescent growth spurt
tended to be similar among sibs of the same family. Thirty three percent of the
variation in peak height velocity and 26% of the variation in peak height velocity
could be attributed to differences between families. Twenty two percent of the
variation in age at take-off could also be so attributed. Unfortunately the standard
errors of these percentages were quite high, so that only peak height velocity reached
twice its SE.
Table 6. Results of the component of variance analysis of growth curve parameters of 105
subjects (boys and girls pooled together after standardisation for sex) distributed among
70 families.
1959). How much of this reduction in variance is due to genes and how much to
environment cannot be estimated from such family data. It reflects not only the
various and complicated interactions of genes and environment, but also the mating
structure of the population, especially of inbreeding.
Table 7. Means and standard deviations (SD) of the standardised Euclidean distances of the
residuals (N = 59)
The mean values are well below zero. comparison of those mean values with the
expected value of zero (under random conditions) revealed significance for all three
combinations of traits. Since the Euclidean distance is a dissimilarity coefficient,
this means that the pattern of residuals of the three considered traits are significantly
more similar to each other than what would be expected if only random effects were
SARSUNA-BARISHA LONGITUDINAL GROWTH STUDY 235
5. DISCUSSION
The Sarsuna-Barisha series has proven to be a very useful source of information to
study variability in human growth patterns. The present overview highlights the
major outcomes of the analysis of this data.
• The charts for growth and growth velocity in height have been useful references
for the growth status of middle class Bengali children but their validity should be
checked in the light of a possible secular trend.
• The analysis of individual growth patterns by means of growth curve fitting has
thrown light on the dynamics of growth in an Indian population. It has shown
that the interrelation between various aspects of the human growth curve and the
dynamics of sexual dimorphism in the growth pattern is quite different between
Indian and Western children.
• The familial structure of the Sarsuna-Barisha data has allowed to examine family
resemblance in various aspects of the pattern of growth, an aspect of the study of
growth which is generally very difficult to tackle because of the lack of proper
material.
• Finally, the comparison of the tiny changes in the patterns growth of three body
dimensions has revealed that short-term variations in growth rate seem to be fairly
synchronised in traits such as height, sitting height and shoulder width.
6. REFERENCES
Boas, F., 1916, On the variety of lines of descent represented in a population. American Anthropologist,
18: 1-9.
Butler, G.E., McKie, M., and Ratcliffe, S.G., 1989, An analysis of the phases of mid-childhood growth
by synchronization of growth spurts. In Auxology 88: Perspectives in the science of growth and
development edited by J.M. Tanner (Niigata-Shi: Nishimur; London: Smith-Gordon), p. 77-84.
236 CHAPTER 18
Cole, T., 1989, Using the LMS method to measure skewness in the NCHS and Dutch national height
standards. Annals of Human Biology, 16: 407-419.
Cole, TJ., and Green, PJ., 1992, Smoothing reference centile curves: the LMS method and penalised
likelihood. Statistics in Medicine, II: 1305-1319.
Das, S.R, 1985, Mixed-longitudinal growth data for 22 measures - The Sarsuna-Barisha series, India.
Volume I boys and Volume 2 girls. Anthropological Survey of India, Calcutta.
Hauspie, R, Das, S.R., Preece, M.A., Tanner, 1.M., and Susanne, e., 1985, Decomposition of sexual
dimorphism in adult size of height, sitting height, shoulder width and hip width in a British and West
Bengal sample. In Human Sexual Dimorphism edited by 1. Ghesquiere, RD. Martin, and F.
Newcombe (Taylor & Francis: London), p. 207-215.
Hauspie, R.C., 1989, Mathematical models for the study of individual growth patterns. Revue
d'Epidemiologie et de Sante Publique, 37: 461-476.
Hauspie, RC., and Chrzastek-Spruch, 1999, Growth models: possibilities and limitations. In Human
Growth in Context edited by F.E. 10hnston and B. Zemel (London: Smith-Gordon) p. 15-24.
Hauspie, Re., and Das, S.R, 1995, Short-term variations in growth rate of height, sitting height and
biacromial diameter in Bengali children. In Essays on Auxology presented to lames Mourilyan
Tanner edited by R Hauspie, G. Lindgren, and F. Falkner (Welwyn Garden City: Castlemead
Publications), p. 260-268.
Hauspie, Re., Bergman, P., Bielicki, T., and Susanne, e., 1994, Genetic variance in the pattern of the
growth curve for height: a longitudinal analysis of male twins. Annals of Human Biology, 21: 347-
362.
Hauspie, Re., Das, S.R, Preece, M.A., and Tanner, 1.M., 1980. A longitudinal study of the growth in
height of boys and girls of West Bengal (India) aged six months to 20 years.
Hauspie, Re., Das, S.R, Preece, M.A., and Tanner, 1.M., 1982, Degree of resemblance of the pattern
of growth among sibs in families of West Bengal (India). Annals of Human Biology, 9: 171-174.
Hauspie, Re., Wachholder, A., Baron, G., Cantraine, F., Susanne, C., Graffar, M., 1980, A
comparative study of the fit of four different functions to longitudinal data of growth in height of
Belgian girls. Annals of Human Biology, 7: 347-358.
10licoeur, P., and Pontier, 1., 1993, Peut-on utiliser des modetes de croissance longitudinale dans
l'analyse de donnees transversales? Biometrie et Praximetrie, 33: 33-44.
Martin, R, 1928, Lehrbuch der Anthropologie. Erster Band-Somatologie (lena: Gustav Verlag).
Meredith, H.V., 1981, An addendum on the presence and absence of a mid-childhood spurt in somatic
dimensions. Annals of Human Biology, 8: 473-476.
Patterson, H.D., 1950, Sampling on successive occasions with partial replacement of units. 10urnal of
the Royal Statistical Society Bulletin, 12: 241-255.
Powell, MJ.D., 1969, Harwell Subroutine Library. Routine VA05A, AERE.
Prader, A., 1982, Biomedical and endocrinological aspects of normal growth and development. In
Human Growth and Development edited by 1. Borms, R Hauspie, A. Sand, e. Susanne, and M.
Hebbelinck (New York: Plenum Press), p. I -22.
Preece, M.A., and Baines, M.l., 1978, A new family of mathematical models describing the human
growth curve. Annals of Human Biology, 5: 1-24.
Smith, C.A.B., 1980, Estimating genetic correlations. Annals of Human Genetics, 43: 265-284.
Tanner, 1.M., 1951, Some notes on the reporting of growth data. Human Biology, 23: 93-159.
Tanner, 1.M., 1959, Boas' contributions to knowledge of human growth and form. In The Anthropology
of Franz Boas: Essays on the Centennial of his Birth edited by W. Goldschmidt. Memoirs of the
American Anthropological Association, nO 89. American Anthropologist, 61: 76-111.
Tanner, 1.M., and Cameron, N., 1980, Investigation of the mid-growth spurt in height, weight and limb
circumferences in single-year velocity data from the London 1966-67 growth survey. Annals of
Human Biology, 7: 565-577.
Tanner, 1.M., Whitehouse, RH., Marubini, E., and Resele, L.F., 1976, The adolescent growth spurt of
boys and girls of the Harpenden Study. Annals of Human Biology, 3: 109-126.
Zemel, B.S., 10hnston, F.E., 1994, Application of the Preece-Baines growth model to cross-sectional
data: problems of validity and interpretation. American 10urnal of Human Biology, 6: 563-570.
CHAPTER 19
S.RAO
1. INTRODUCTION
237
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 237-250.
© 2001 Kluwer Academic Publishers.
238 CHAPTER 19
life. The present paper therefore, examines the effect of undernutrition on the growth
spurt as well as on other linear components of growth.
2.1 Subjects
A longitudinal adolescent growth study was undertaken in seven villages around
Pune during 1992-96. All children in the age group 9-18 year from these villages
were considered for the study. Most children were contacted in school while
remaining were recruited through house-to-house visit and an attempt was made to
contact a maximum number of study children in each subsequent round in order to
reduce the loss in follow-up. The majority (79%) of parents had cultivation as major
occupation, but most were engaged in subsistence farming due to poor land holdings.
People generally consumed vegetarian diets consisting of cereals, pulses and
vegetables, and the consumption of milk, meat, fruits was negligible. None of the
villages had safe drinking water supplies and lacked adequate hygiene and sanitary
conditions. Body weight, height, sitting height, foot length, shoulder width were
measured by a trained team of researchers once in every six months.
2.2 Measurements
Measurements were considered one year (±1 week) apart and all available
observations were used in the analysis. Weight (WT) was measured up to 20 g using
an electronic weighing balance (ATCO, India) while height (HT) was measured up to
0.1 cm using stadiometer (UNA, India). The same equipment was used to measure
sitting height (SHT), while leg length (LEGLN) was estimated by subtracting
sitting height from stature. Foot length (FLN) was measured with the help of a
specially designed board fixed with a fine metal ruler while shoulder width (SW) was
measured with the help of a measuring tape. All the measurements were recorded by
trained investigators.
(Figure 2) by age. The velocity curve for height shows a shift to the right as
compared to the NCHS velocity curve for healthy children. The maximum
increments in weight (4.65 kg/year) and height (6.4 cm/year) occurred around the age
of 14+ years, which is higher than that reported for well-to-do Indian children
(Vijayaraghavan, Gowrinath and Indubhushan 1974).
70l • P50
180
.• ' P50
Weight 170J Height
60
160
50
eo 8u 150
~
•....40
..<::
eo
'Q) ~140
0)
~30 ::t:
130
20
120
10 110
01 100
8 10 12 14 16 18 20 8 10 12 14 16 18 20
Age, years Age, years
Figure 1. Mean weight and height of rural boys by chronological age compared to the 3rd
and 50th NCHS centiles
Table 1. Mean and SD afweight and height· rural boys
weight, kg height, cm
Age, years n mean SD mean SD
8+ 86 18.98 ** 1.9 118.2 ** 5.5
9+ 190 20.60 ** 2.3 122.8 ** 5.3
10+ 259 22.59 ** 2.7 127.7 ** 5.6
11+ 252 24.70 ** 3.5 132.0 ** 6.3
12+ 208 27.65 ** 4.6 137.4 ** 7.5
13+ 193 30.37 ** 6.6 142.1 ** 8.3
14+ 195 33.46 ** 6.1 147.7 ** 8.4
15+ 163 37.37 ** 6.5 152.7 ** 8.0
16+ 122 40.68 ** 6.1 157.0 ** 7.9
17+ 76 43.74 ** 6.0 160.0 * 7.3
18+ 33 45.59 5.8 161.2 7.5
19+ 8 50.24 7.0 164.7 5.3
*: p < .05 **: p < .01
240 CHAPTER 19
The difference is not so much in actual peak height velocity, but the age at which
peak height velocity occurs is significantly (p < 0.05) different. Adolescent growth
of rural boys is thus delayed by 1-2 years in comparison to healthy children. The
velocity curves also show higher height gains compared to healthy boys in later
ages. Thus, slow but gradual height gains as late as 18+ and 19+ years not only
prolong the adolescent growth span, but are indicative of catch-up growth in rural
Indian children.
7 8
Weight velocity Height velocity
6 7
a
Q)
a6
<-OIl 5 ~
..:.: ] 5
>. 4 >.
., 'g
.<;::
u
0 4
;:-
3 ~
.:E00
'0)
~3
'0)
:=: 2 ::r::
2
'P50
01 " o 1 "
8 10 12 14 16 18 20 8 10 12 14 16 18 20
Age, years Age, years
Figure 2. Weight and height velocity by chronological age for rural boys compared with
the NCHS PSO
Table 2. Attained weight and height for normal ("N") and underweight ("U") boys. All
differences (Dif) were statistically significant (p < .01)
8+ 23 63 21.3 1.3 18.1 1.2 3.2 123.4 4.1 116.4 4.7 7.0
9+ 45 145 23.6 1.6 19.7 1.7 3.9 128.1 3.8 121.2 4.6 6.9
10+ 59 200 26.0 2.4 21.6 1.8 4.4 133.6 4.4 125.9 4.7 7.7
11+ 52 200 29.1 3.5 23.6 2.4 5.6 139.1 4.6 130.2 5.4 8.9
12+ 38 170 34.0 4.8 26.2 3.1 7.8 146.5 5.8 134.5 6.2 12.0
13+ 23 170 39.7 3.1 28.8 4.3 10.9 154.8 4.1 140.1 7.1 14.4
14+ 21 174 44.0 3.1 32.2 5.1 11.8 159.4 4.4 146.3 7.6 13.1
15+ 19 144 48.2 2.8 35.9 5.4 12.3 162.4 4.4 151.4 7.5 11.0
16+ 12 110 51.2 2.6 39.5 5.2 11.7 165.5 5.0 156.1 7.7 9.4
17+ 9 67 54.4 4.1 42.3 4.6 12.1 168.2 3.5 159.0 7.0 9.2
18+ 4 37 57.9 6.0 45.3 5.0 12.7 170.6 3.7 160.9 6.9 9.1
The velocity curves based on cumulative data on mean annual increments for
these groups also bring out the delay in adolescent growth among underweight as
well as stunted boys (Figure 3) as compared to their normal counterparts. The
differences in maximum increments either for weight or height, were not prominent
but were significant for ages at which maximum increments occurred. Thus, more
than growth rates, it is the timing of peak height (or weight) velocity which is more
sensitive to undernutrition. This perhaps results in significant gains at later ages
among undernourished boys compared to normals.
1997)0 However, no such data, other than weight or height growth, is reported on
children from rural Indian communities wherein undernutrition has been a persistent
problemo
7 7
Weight velocity Weight velocity
6 6
...til Normal ...til
~ 5 ~ 5
---
00
.>d
---
00
.>d
o~
u
4 o~
u
4
0 0
~ ~
;;- ;;-
Q
3 3
.:E00 .:E00
0d) 0 0d)
~ 2 ~ 2
01 0
8 10 12 14 16 18 20 8 10 12 14 16 18 20
Age, years Age, years
91 9
Height velocity Height velocity
8 8
8u 6
0 8u 6
>. 5
0<;::: o~ 5
0
u .. u
0
~ 0 ~
;;- 4 ;;- 4
.:E00 .:E00
0d) 0d) 3
3
/--
::t: ::t:
2 0 2
O~I---.---.--~--~--~--. O~I---.---.--~--~--~--.
8 10 12 14 16 18 20 8 10 12 14 16 18 20
Age, years Age, years
Figure 30 Weight and height velocities by chronological age for rural boys
GROWTH IN RURAL INDIAN CHILDREN 243
Table 3. Mean and standard deviation SD for sitting height (SHT) and leg length (LEGLN)
of boys and girls.
Boys Girls
sitting leg length, sitting leg length,
Age height, em em height, em em
yrs n mean SD mean SD n mean SD mean SD
9+ 64 65.5 2.8 59.0 3.9 66 64.8 2.8 59.2 3.7
10+ 120 67.2 2.9 61.9 3.8 125 66.9 2.9 62.0 3.9
11+ 204 68.7 3.2 64.4 4.1 191 69.1 3.5 65.2 4.0
12+ 267 70.3 3.2 67.2 3.8 202 71.8* 3.5 67.8 3.7
13+ 321 72.8 3.7 70.4 4.4 204 74.0* 4.0 70.2 3.8
14+ 315 76.1 4.6 73.6 4.8 209 76.5 3.6 71.8* 3.8
15+ 270 79.3 4.5 76.3 4.5 157 78.0* 3.3 72.6* 3.7
16+ 210 81.5 4.4 77.8 4.6 125 78.4* 3.3 72.5* 3.8
17+ 130 83.1 4.0 78.8 4.5 95 79.0* 3.1 73.0* 4.2
18+ 61 84.3 3.6 79.5 4.2 36 78.1* 2.5 72.3* 4.2
19+ 23 85.2 2.6 79.7 4.8 11 80.3* 2.2 71.5* 3.0
*: p < .05
The cumulative data on annual increments was used to develop velocity curves
(Figure 4). It was observed that mean annual increments in sitting height were
smaller than mean annual increments in leg length up to 13 years in boys and 10
years in girls. However, velocities in sitting height became significantly (p < 0.01)
greater than velocities in leg length beyond 14 years in boys and 11 years in girls. It
was thus observed that leg length grows faster and reaches its peak velocity before
sitting height, i.e. at 13+ years and 14+ years respectively for boys and at 10+ years
and 11 + years respectively for girls. In fact, it showed that growth of leg length in
relation to stature growth was faster before spurt among girls than boys.
8
Boys 81 Girls
7 7
6 6
a ....
Q)
'"
Q)
~ 5 ~ 5
u u
z:; 4 SHT z:; 4
·u0 ·u0
~
"i3 3 ~ 3
>- D' '.
• 0
2 o--o-G ... 2
oI , , , , , , 0
8 10 12 14 16 18 20 8 10 12 14 16 18 20
Age, years Age, years
Figure 4a. Velocity curves for height (RT), sitting height (SRT) and leg length (LEGLN)
in boys and girls
8l Boys 81 Girls
7 7
6 6
....
~
....
~
Q) Q)
~5 ~5
u u
z:; 4 z:; 4
'uo '0
0
"i3 3 "i3 3
>- >-
2 2
0- ,,_,,_0 _,,F~N
~-G_
~ '1 0- "-G_ FLN
°-0_
oI , , , ,0-,-0 , 0 , , ,0- 0 -,-
8 10 12 14 16 18 20 8 10 12 14 16 18 20
Age, years Age, years
Figure 4b. Velocity curves for height (RT), foot length (FLN) and shoulder width (SW) in
boys and girls
GROWTH IN RURAL INDIAN CHILDREN 245
Table 4. Mean and standard deviation, SD, for foot length (FLN) and shoulder width (SW)
of boys and girls.
Boy"s Girls
foot length, shoulder width, foot length, shoulder width,
Age em em em em
Velocity curves for these measurements show that mean velocities in foot length
were larger before peak height velocity but taper off gradually both in case of girls
and boys (Figure 4). The age at maximum velocity of foot length was 13+ years in
boys and 10+ years in girls, i.e. one year before their age at peak height velocity.
Growth in foot length was faster before peak height velocity among girls than boys.
The trend in growth of foot length was thus almost similar to that observed in leg
length.
The age at maximum increment for shoulder width was however, 15+ years in
boys and 13+ years in girls. Thus, shoulder width reaches its peak value two years
after peak height velocity and one year after sitting height reaches its peak, both in
case of girls and boys. The trend in growth of shoulder width was thus consistent
with the growth in sitting height.
In general, it was observed that age at maximum increment for each measurement
was earlier for girls than boys. Further, the sequence of attainment of maximum
growth in these measurements was much the same in boys and girls, but the ages at
which maximum growth is achieved differed. Thus, the spurt in foot length and leg
length occurred before the spurt in height velocity while that for sitting height and
shoulder width occurred after peak height velocity in both sexes.
Sitting height and leg length of rural boys is compared with reported Indian
studies (Agarwal, Agarwal, Upadhyay, Mittal, Prakash and Rai 1992, Dasgupta and
Das 1997) and British data (Eleveth and Tanner 1976) in Figure 5. It shows that rural
boys have significantly lower leg lengths. However, this difference is reduced in case
of middle-class Bengali boys while Indian well-off have in fact higher leg lengths
than British children. Sitting height too showed a similar socio-economic gradient,
246 CHAPTER 19
but, unlike leg length, even for well-off Indian boys, it was lower throughout
adolescence when compared to British children. In fact, the difference in final size for
leg length of rural boys, when compared with British children, was only 3 cm (79.3
Vs 82.9 cm) but was 8.4 cm (83.9 Vs 92.3 cm) in the case of sitting height.
110 110
100 Sitting height 100 J Leg length
§ Indian well-off British 81 Indian well-off
.:E 90 90
.-~ 80
u.
~ £
~
5 80
gf ~ ~. British
.-
1 •
'B 70 OJ
~
70
(/)
60 60
50 50
9 10 11 12 13 14 15 16 17 18 9 10 11 12 13 14 15 16 17 18
Age, years Age, years
110 110
100 Sitting height 100 Leg length
§ Middle class British 8 Middle class
..... 90 Bengal u 90 Bengal ..
.-~ 80
~ -5
5
~
BntIsh
~
~
-
~
80
:€
(/)
70 ~ 70
60 60
50 50
9 10 11 12 13 14 15 16 17 18 9 10 11 12 13 14 15 16 17 18
Age, years Age, years
110 110
100 Sitting height 100 I Leg length
§ Indian rural British 8 1 Indian rural
,E. 90 ... -- u. 90 B't'h
.-~
1
~ £ n IS
-
5
~
80 80
~
.-
~ ~
'B 70 OJ
~
70
(/)
60 60
50 50
9 10 11 12 13 14 15 16 17 18 9 10 11 12 13 14 15 16 17 18
Age, years Age, years
Figure 5. Comparison of sitting height and leg length of Indian well-off, Bengali middle
class and Indian rural boys with British boys.
GROWTH IN RURAL INDIAN CHILDREN 247
Similar comparisons of observed mean foot length and shoulder width values of rural
children with British children showed that foot length values were significantly (p <
0.01) lower throughout adolescence and the differences increased progressively up to
adulthood. However, differences in shoulder width were not prominent when
compared with British children.
Rural Indian children are known to suffer from malnutrition in early life, as also
from a lack of optimal nutrition during adolescence. It would be therefore, worth to
examine ages at maximum increment for various physical dimensions and compare
with well-to-do Indian children for examining the effect of undernutrition on linear
components of growth.
We have observed in the previous section that differences in sitting height were
more prominent than those in leg length in case of rural children. Thus,
undernutrition prevailing in rural communities has affected all the components of
linear growth but sitting height is relatively more affected than other parts.
The rural group as a whole, however, consists of individuals suffering from both
short-term and long-term undernutrition. Therefore, in order to examine the effect of
current undernutrition, children were classified as normal (weight-for-age ~ 75) or
undernourished (weight-for-age < 75) as before and their body measurements were
compared.
Underweight children had lower attained values for all the measurements viz., leg
length, sitting height, foot length and shoulder width compared to normal children
(Figure 6). Further, it can be observed (Figure 6) that children with normal weight
status had leg lengths similar to that of British children, but had significantly lower
sitting heights. Therefore, most of the difference in adult height was due to
differences in sitting height. However, underweight children had lower values for
sitting height and leg length when compared with their normal counterpart as well as
British children. Thus, differences in adult height were mainly attributed to
differences in sitting height in case of normal children, but additionally to differences
in leg length in the case of underweight children.
Similar comparisons for foot length and shoulder width showed (Figure 6) that
mean foot length values for normal boys were close to British values up to 15 years
of age, but were lower thereafter. In contrast, mean values for shoulder width for
normal boys were higher than British values throughout adolescence. Thus, children
with normal weight status during adolescence had satisfactory leg length, foot length
and shoulder width, but lower sitting height while underweight children had lower
values for all body dimensions.
Undernutrition delayed ages at maximum increments of all the measurements, but
the sequence of their attainment remained the same as in normal children. Despite the
fact that for underweight boys, velocities at later ages were higher for all the
measurements, compared to normal children, their final adult sizes were lower than
the normal children. Thus, although there is some evidence for catch-up growth, it
was not sufficient to achieve comparable final adult sizes.
248 CHAPTER 19
4. SUMMARY
In conclusion, rural children had lower attained values than British or Indian well-
off children for all the linear components of stature growth considered in this study.
The sequence of the adolescent spurt in these linear components is the same in boys
as in girls, i.e. the peak velocity in foot length and leg length is reached before peak
height velocity, while that for sitting height and shoulder width occurs after peak
height velocity. While foot length was the first trait to reach peak velocity, shoulder
width was the last one in this sequence. The ages at maximum increment for each of
the linear components were earlier for girls than boys.
100, 100
Sitting height Leg length
I
au 90 90
au
~80 0;9" 80
'i) 01)
..c: I':
(!.)
.S
t::
70 -; 70
(!.)
....l
i;i)
60 • Normal
---{]-- underweighl
60 ~ • Normal
---{]-- Underweight
.. ------- British - - - - - - - - - British
501 , , , , , , , , , , 50
9 10111213141516171819 910111213141516171819
Age, years Age, years
28, 45
Foot length I Shoulder width
26
a § 40
u 0;9"
0;9" 24 -0
01)
I': .~ 35
~ (!.)
-0
-00 22 "3
0
t5
~
30
____ Normal
---{]-- Underweight
~-
• Normal
---{]-- Underweigh
--------- British --------- British
181 , , , , , , , , , , 25
9 10111213141516171819 910111213141516171819
Age, years Age, years
Figure 6. Comparison of British, normal and underweight boys for sitting height, leg
length, foot length, and shoulder width.
The growth of length in relation to stature growth was faster before the spurt
among girls than among boys. In general, trends in growth of foot length and
shoulder width were similar to those observed for leg length and sitting height
respectively.
GROWTH IN RURAL INDIAN CHILDREN 249
Undernutrition during adolescence not only delayed the adolescent growth spurt,
but also affected all the linear components of growth. Comparison of rural data with
Indian well-off and British children revealed that the impact of undernutrition was
relatively more on sitting height as compared to other measurements. Further, it was
observed that leg length improved as social class improved, and Indian well-off
children had actually higher values of leg length than British children, but this was
not true for sitting height. It thus, appeared that growth in leg length was influenced
by post-natal environmental influences while growth in sitting height had probably
influence of intra-uterine nutritional environment apart from genetic influence. In
that case, relative growth in sitting height to leg length can provide an index of
disproportionate growth.
Recent hypothesis of 'foetal origins of adult diseases' describes risks of adult
diseases in relation to disproportionate growth at birth (Barker 1994). In view of
this, the findings presented in this paper indicate the need for more observational
studies to understand implications of disproportionate growth of linear components
in adulthood in relation to the risk for adult diseases. The likelihood for such a
relationship appears possible as it is known that prevalence of hypertension, diabetes
and coronary heart disease are low in rural populations and increase as we go to urban
middle and urban affluent classes. The observations discussed in this paper thus have
significant implications for adult health.
5. REFERENCES
Agarwal, D.K, Agarwal., KN., Upadhyay, S.K., Mital, R, Prakash, R, and Rai, S., 1992, Physical and
sexual growth pattern of affluent children from 5 to 18 yrs of age. Indian Pediatrics, 29, 1203-1283.
Banik, N.D.D., 1982, Semi-longitudinal growth evaluation of children from birth to 14 yr in different
socio-economic groups. Indian Journal of Pediatrics, 19,353-359.
Barker, DJ.P., 1994, Mothers, Babies and Diseases in Later Life (BMJ Publishing Group).
Behrman, R.E., Kleigman, R.M., Nelson, W.E., and Vaughan, V.C., 1992, Textbook of Pediatrics
(Philadelphia, P.A.W.B.: Saunders Company), 14th Edition.
Dasgupta, P., and Das S.R., 1997, A cross-sectional growth study of trunk and limb segments of the
Bengali boys of Calcutta. Annals of Human Biology, 24, 363-369.
Eveleth, P.B., and Tanner, J.M., 1976, Worldwide variation in Human Growth (Cambridge University
Press: Cambridge).
Pathmanathan, c., and Prakash, S., 1994, Growth of sitting height, subischial leg length and weight in
well-off North-Western Indian children. Annals of Human Biology, 21, 325-334.
Satake, T., Kikuta, F., and Ozaki, T., 1993, Ages at peak velocities and peak velocities for seven body
dimensions in Japanese children. Annals of Human Biology, 20, 67-70.
Tanner, J.M., Hayashi, T., Preece, M.A., and Cameron, N., 1982, Increase in length of leg relative to
trunk in Japanese children and adults from 1957-1977: Comparison with British and Japanese
Americans. Annals of Human Biology, 9, 411-423.
Tripathi, A.M., Agarwal, D.K., and Agarwal, KN., 1976, Physical growth during adolescence in Delhi
school children. Indian Pediatrics, 13, 191-200.
Verghese, KP., Scott, RB., Teixeira, G., and Ferguson, A.D., 1969, Studies in growth and development,
XII. Physical growth of North American Negro Children. Pediatrics, 44, 243-249.
250 CHAPTER 19
Vijayaraghavan, K., Gowrinath, S.T., and Indubhushan, M.e., 1974, Growth performance of well-to-do
Hyderabad children - a follow up study. Indian Journal of Medical Research, 62, 117-124.
CHAPTER 20
T. MOFFAT
1. INTRODUCTION
transportation and communication lines. Studies that have been done in Nepal in
various locations document an extremely high prevalence of stunting for children
less than five years of age with little change in growth status from the 1970s to the
1990s (Brink et al. 1976, Martorell et al. 1987, Costello 1989, UNICEF 1996,
Panter-Brick 1997).
Much less research on child growth has been conducted among urban children in
Nepal. An exception here is Panter-Brick et al. /s (1996) study of the growth of
homeless urban boys compared to urban rural boys. Nepal is a mainly rural, but
rapidly urbani sing country; in the past two decades it has undergone an unprecedented
rate of urbanisation (UNICEF 1992). The studies that do examine urban children
assume a rural/urban contrast without exploring the complex mosaic of differing
socio-economic landscapes within urban environments (Brockeroff 1995). Thus,
intensive research on urban sub-populations is needed to understand the complexity
of child growth within the larger urban communities.
Those living in peri-urban locations are noted as being distinct within the urban
environment. Peri-urban communities are situated on the periphery of the city and
are comprised mainly of rural-to-urban migrants working in industry or living in
squatter settlements. In general, people living in peri-urban communities are more
impoverished than those living within the city proper having less resources and
essential services such as electricity, piped water and sewage systems (Harpham et al.
1988).
This study evaluates the growth of a sample of children less than five years of
age residing in peri-urban Kathmandu. Growth is examined by employing both
cross-sectional and longitudinal anthropometric data in order to get at what is
described above as both the process and the end product of linear growth retardation.
Aspects of growth faltering in low-income countries such as the relationship
between wasting and stunting, gender differences, seasonality, and comparisons to
children in rural locations are examined.
2. METHODS
the cold season. Heavy sweaters and belts, however, were removed. Children's weight
was measured with a portable suspended infant/child weighing scale (Perspective
Enterprises, PE-HS-25, Kalamazoo, MI) to the nearest 0.1 kg. The scale was set to
zero at the beginning of each session and checked periodically throughout the
sessions. The scales were calibrated bi-monthly with a 2 kg weight and there were no
changes detected throughout the study period. Recumbent length was taken for
children up to and including 24 months, and height was measured for those over 24
months with a portable wooden adult/infant measuring board (Perspective
Enterprises, PE-AIM-lOl, Kalamazoo, MI) to the nearest millimetre. Arm
circumferences were measured to the nearest millimetre with insertion slot measuring
tapes (Perspective Enterprises, Kalamazoo, MI) and triceps skinfolds in millimetres
with a Lange skinfold calliper (Beta Technology Incorporated, Cambridge, MD). For
triceps skinfolds, three measurements were taken in quick succession and an average
of the three was used as the measurement (Frisancho 1990).
Nepalese children have no official birth certificates, so the age of the child had to
be ascertained from the parents. In most cases parents did not know the year of birth,
but they did know the age of the child and the month and day on which the child was
born. Dates were converted from the Nepalese calendar to the Western calendar and
then checked to confirm that they corresponded with the age stated by the parent. The
caregiver was asked at each subsequent measuring session to state the date of birth
and the age of the child in order to check the accuracy of the information.
3. RESULTS
Table 1. Mean and standard deviation (SD) of heightf and body weight of children aged 0-
60 months, measured at first survey
A comparison of mean HAZ, WAZ and WHZ scores by age for both sexes
presented in Figure 1 shows that, apart from the infants, these children, on average,
are moderately-to-severely retarded in terms of linear growth and mild-to-moderately
underweight for age. Their WHZ scores, a measure of acute weight loss, however,
are very close to the NCHS reference median, indicating that they are not wasted
relative to their body frame.
GROWTH STUNTING IN NEPALESE CHILDREN 255
Table 2. Mean arm circumference (e), triceps skinfoids (Ts) and Upper Arm Areas(UMA)
of children aged 12 - 60 months, measured at first survey.
Table 3. Mean and standard deviation (SD) ofZ-scoresfor HAZ, WAZ and WHZ in males
and females, aged 0 - 60 months, at first survey.
o
-0.5
- I f- ........................ ~
~ -1 .5 f- .........................= ...................
~
8
N -2 f- ............................................ ······~··················· f
VJ
-3.5 -=--=-=----:----:--------------~
o - 5.9
,-I
Figure 1. Mean Z-scores of HAZ, WAZ and WHZ by age category for all children measured
for the first time (n = 283)
Figure 1 shows the progressive decrease in mean Z-scores for all three indices as
the ages of the children increase. One-way ANOV A tests of all indices indicate that
the mean Z-scores for children in each age category are significantly different (p <
0.0001). A Tukey-HSD post-hoc analysis of the means indicates that for HAZ,
children less than six months old are significantly different from all other age
groups; children from six to 12 months are significantly different from two, three
and four-year olds and one-year olds are significantly different from two, three and
four-year olds. For WAZ, only infants « 12 months) are significantly different from
all other age categories. For WHZ, all age groups differ significantly from the one-
year olds. This analysis indicates that under-weight-for-age and stunting peaks during
years one and two respectively, and continues on through the fourth year of life. The
most serious state of low WHZ, or wasting, occurs among the one-year olds.
Upper arm circumferences were converted to muscle areas and then converted to
Z-scores (ZAM) along with triceps skinfolds (ZTR) relative to the NCHS reference
medians (Frisancho 1990) (Table 4). Mid upper arm circumference Z-scores (ZCIRC)
for all children range between -2 and -1 SD below the NCHS reference medians.
Triceps skin fold Z-scores (ZTR), however, are all close to the reference median, i.e.
they are abnormal or slightly above normal in terms of fatness for that area (Figure
2). The arm measurements corroborate what was found for height and weight; that is,
these children are small but not wasted. Low arm muscle area Z-scores (ZAM)
indicate that there is more wasting of the upper arm muscle than there is of upper
arm fat. It has been proposed that arm circumference be used as a quick and simple
measure of malnutrition (Anderson 1979). Arm circumference and arm muscle area,
however, do differ as shown by their respective Z-scores, ZCIRC and ZAM (Table
4). A child can appear to be better off because he/she is plump and has a moderate
GROWTH STUNTING IN NEPALESE CHILDREN 257
size arm circumference, but the child may still have poor muscularity as indicated by
a low ZAM.
Table 4. Mean and standard deviation (SD) ofZ-scoresfor triceps skinfolds (ZTR), mid
upper arm circumference (ZCIRC) and upper arm muscle area (ZAM) of children aged I2 -
60 months.
0 ZAM
0.51··~ ZTR
0
~ -0.5
0
u
N'" -1
I
-1.5
-2
-2.5
12-23.9 24-35.9 36-47.9 48-59.9
Age, months
Figure 2. Mean Z-score for upper arm muscle area (ZAM) and triceps skinfolds (ZTR) for
all children measuredfor the first time (n = 194)
258 CHAPTER 20
One-way ANOV As for mean ZCIRC, ZTR and ZAM by age category indicate
that ZCIRC does not differ significantly in terms of age. This conforms to the
generalisation that upper arm circumference can be used as a test of malnutrition
independent of age, because it remains relatively constant from 12 to 60 months of
age (Anderson 1979, Gorstein et al. 1994). Mean ZTRs do differ significantly by age
category (p = 0.00). A post-hoc Tukey's-HSD test at the significance level of 0.05
indicates that one year olds have a significantly lower mean ZTR than the older age
groups. ZAM, which takes both ZCIRC and ZTR into account, is also significantly
different by age category (p = 0.01); three and four-year-olds have significantly lower
ZAM scores than one and two year olds. These age trends in arm measurements
corroborate height and weight age trends. The lowest skinfold thickness occurs
among one year olds, which corresponds with the period of greatest wasting or low
weight-for-height. The lowest mean ZAMs occur after the first year of Hfe, mirroring
the drop in mean W AZ and HAZ. After two years of age children remain short, with
low muscularity, but not thin.
Bivariate correlations between ZAM and HAZ are positive and significant (r =
0.32, p = 0.00), but only moderately so. ZAM and W AZ are more highly positively
correlated (r = 0.47, p = 0.00). Thus, children who are stunted and underweight are
likely to have poorer muscularity and vice versa. ZTR is also positively correlated
with WHZ (r = 0.47, p = 0.00). In other words, on average, those who are wasted
will have correspondingly low triceps skinfold scores. There is no significant
correlation between ZTR and HAZ (r = 0.-14, p = 0.05) and a weak one between
ZTR and W AZ (r = 0.22, p = 0.002). This makes sense because both ZTR and WHZ
are measures of thinness, whereas W AZ and HAZ are measures of body size.
3.2 Prevalence of stunting, wasting and low weight among infants and children
Since many studies report anthropometric findings in terms of prevalence of
stunting, underweight and wasting (HAZ, W AZ and WHZ < -2 SD), it is important
to report these results here. Moreover, although the mean Z-scores may show where
the group as a whole stands, they do not indicate what proportion of the sample is in
the lower range of the NCHS reference population.
Table 5. Prevalence of low (5 - 2SD) ht/age (stunted), wt/age (underweight) and wtlht
(wasted) for children aged 0 - 60 months
70
60
50
1
p...
40
30
________ peri-urban
20 ~ rural national
--i'r- urban national
10
6 - 1l.9 12 - 17.9 6 - 36.9
Age, months
Figure 3. Stunting (HAZ < -2SD) prevalence by age category of Nepalese National Survey
of urban and rural population (UNICEF 1996) and peri-urban sample.
18
16
----0-- US 3rd
14 ~US50th
12 - Nepalese boys
~'" 10
0
E 8
\0
] 6
4
2
0
0 6 12 18 24 30 36 42 48 54 60
Age at second measurement, months
Figure 4. Linear growth velocity (Round I to Ill) for Nepalese boys relative to US
longitudinal data (3rd and 50th percentiles)
18
16
----0-- US 3rd
14 ~US50th
] 6
4
2
0
0 6 12 18 24 30 36 42 48 54 60
Age at second measurement, months
Figure 5. Linear growth velocity (Round I to Ill) for Nepalese girls relative to US
longitudinal data (3rd and 50th percentiles)
Mean linear growth rates from Table 6 (Rounds I to III) are plotted against
normative data (US 3rd and 50th percentiles) from Fe1s longitudinal study
GROWTH STUNTING IN NEPALESE CHlLDREN 261
(Baumgartner et at. 1986) for linear growth velocities (cml6 months) (Figure 4). In
this sample Nepalese males have a slow linear growth rate compared to US males;
the Nepalese means track the 3rd percentile until the third year of life, after which
they approach the 50th percentile. On the other hand, the mean linear growth rate for
females declines precipitously from 6 to 18 months, passing from the 50th down to
the 3rd US percentile. As with the males, however, the mean linear growth rate
increases after the third year of life, approaching the 50th percentile again.
Table 6. Mean and standard deviation (SD) of growth rates (cml6 mths) of stature and
body mass for children aged 6 to 60 months between Rounds I and Ill.
Table 7. Repeated measures ANOVA tests for differences in mean HAZ taken for children
measured in Rounds I, II and III and covaried by age at measurement (n = 107)
not a significant covariate for WHZ, it was not included in the repeated measures
ANOVA test for differences in mean WHZ by season.
As can be seen from the results of the repeated measures ANOVA tests (Tables 7
to 9), there are no statistically significant differences in HAZ and W AZ in cold, hot
and rainy seasons. Mean WHZ, however, does vary by season (Table 9). Mean WHZ
scores for children measured in the winter, hot and rainy seasons respectively are as
follows: -0.40, -0.72, -0.84.
Table 8. Repeated measures ANOVA tests for differences in mean WAZ taken for children
measured in Rounds I, II and III and covaried by age at measurement (n = 113)
Table 9. Repeated measures ANOVA tests for differences in mean WHZ taken for children
measured in Rounds I, II and III (n = /07)
4. DISCUSSION
Linear growth retardation or stunting is one of the most prevalent forms of growth
faltering affecting children in low-income countries. It is estimated that
approximately 50 to 70% of pre-school children in South Asia are growth stunted
(Victora 1992). This is also the case in Nepal (UNICEF 1996), as well as for the
peri-urban sample reported on here.
Although linear growth retardation, or stunting, is the salient issue for these
children, there may be some connection between growth in stature and body mass.
Waterlow (1992) is adamant that linear growth retardation and wasting are separate
physiological processes, although he has since acknowledged that there is evidence
that episodes of wasting may precede stunting (Waterlow 1994). Naborro et al.
(1988), Costello (1989) and Panter-Brick (1997) all show evidence that the thinnest
children in their samples gained relatively more body weight than other children in
succeeding months, but at the expense of growth in stature. In terms of cross-
sectional growth for this peri-urban sample, this relationship can be seen from the
differential age pattern of growth status. The lowest mean WHZ occurs in the sample
from 12 to 24 months of age. This thinness is corroborated by a significantly lower
GROWTH STUNTING IN NEPALESE CHILDREN 263
mean ZTR for this age group. Notably, this state precedes the peak in low mean
HAZ, or stunting, from 24 to 60 months.
Longitudinal growth patterns lend a more dynamic view of child growth.
Although cross-sectional or attained growth indicates that the Nepalese children in
this sample were not stunted until after 12 months of life, boys from early infancy
have low linear mean growth velocities. Girls' mean velocity, however, only drops
after the first year. As mentioned above, this may be due to the latter being
significantly smaller at birth. Lack of birth weight data is a drawback of this study,
as there is evidence that birth weight may determine the pace and absolute size of
early infant growth (Chen et al. 1980). Early infant growth faltering, moreover, may
be regulated by deficient in utero nutrition and poor micro-nutrient stores at birth
rather than the postnatal environment (Allen 1994).
Linear growth velocities do improve after the third year of life, when both males
and females approach the 50th percentile of the US normative data. This is in
contrast to cross-sectional attained growth where three and four-year-olds have the
lowest mean HAZ scores and the highest prevalence of stunting. The difference in
these two types of measures illustrates the distinction between the process of
stunting and the state of being stunted. The process of stunting began at six months
(3 months for boys) and lasted through to 36 months, but the children were stunted
from 24 months onward. There appears to be an increase in linear growth velocity
for both male and female three and four year olds. This corresponds with what is
termed catch-up growth for three and four year old children in low-income countries
(Martorell et al. 1994), although it is not enough to overcome original deficits.
This state of being stunted is also corroborated by significantly reduced upper arm
muscularity for three and four-year-olds compared to the younger children. Results
from arm measurements indicate that there is more wasting for these children in
upper arm muscle compared to mid-upper arm body fat. This is further evidence of
protein-energy-malnutrition; children may be obtaining enough calories, but not
enough protein to build body muscle (Frisancho 1990).
There appear to be some gender differences in arm measurements. Interestingly,
among the four-year-olds (Table 4), it is the girls who have slightly higher mean
ZAM scores, whereas among the three-year-olds, the boys have higher mean ZAM
scores. This may be due to cohort differences rather than systematic gender
differences. Indeed, there is no difference in mean ZAM scores between boys and girls
for the sample as a whole.
The only difference in mean Z-scores between the sexes is found for mean HAZ (t
= 2.74, df = 38, p = 0.009) and W AZ (t = 2.91, df = 38, p = 0.006) of infants less
than six months of age (Table 3). In this case the mean Z-score is lower for females
than males. It is surprising that female infants would be significantly smaller than
male infants at such a young age, as gender differences in Z-scores usually signal
preferential care of boys over girls. Infants less than six months are being breast-fed
with little in the way of supplementation, so they would not have much opportunity
for preferential treatment through better food. It is possible that mothers are breast-
feeding male infants more frequently than female ones, although it is difficult to
prove this without detailed observational data.
While females begin growth faltering after six months of age, males appear to
have slower growth rates earlier in the infancy period (Figures 4 and 5). The females
less than six months of age, as discussed above, are significantly smaller than their
264 CHAPTER 20
male counterparts. Therefore, the faster female infant growth velocity may be due to
post-natal (in the first six months) catch-up growth and regression to mean. See
Baumgartner et at. (1986) for explanations of these phenomena. That is, they may
have been growing faster during the infancy period to make up for deficits in size at
birth. This may explain why they are comparable in size to the males in later years.
Unfortunately, without birth weights, it is difficult to test this hypothesis. Birth
weights were unobtainable for this sample because more than half of the children
were born at home and no birth weight was recorded. Those born in the hospital
were not given a record of their birth weight.
There is no doubt that discrimination against girl children is an important issue
in South Asia and evidence of preferential care of males over females has been
highlighted in studies of household food allocation in Nepal; however, differences are
found only during adolescent and adult years (Gittelsohn et at. 1997). The fact that
no systematic evidence of gender differences in body size was found in this peri-urban
sample concurs with the results of other growth studies in Nepal (Panter-Brick 1997)
and Himalayan India (Himmelgreen et at. 1991). Himmelgreen et at. (1991), in fact,
argue that the assumption of female nutritional disadvantage in northern India should
be re-evaluated given the lack of gender differences in their findings. There is
evidence from an Andean community in Peru that infants and pre-school children of
both genders are buffered by nutritional stress due to child preferential food allocation
within the household (Leonard 1991). Thus, gender differences may be more
pertinent to adolescent age groups than infants and pre-school children.
An investigation of urban-rural differences in Nepal is made in comparing the
results from this peri-urban sample to those from the 1995 Nepalese National Health
Survey (UNICEF 1996). The children in this peri-urban sample, at least for those
over 12 months of age, have a prevalence of stunting in the same range as urban
Nepalese children; both of these groups have a lower prevalence of stunting than the
rural Nepalese children. This points to the possibility of slightly better linear growth
status for the peri-urban children, a result which is often found for comparisons
between urban and rural children in other low-income countries (WHO 1988,
Williams 1990).
One way that growth retardation may manifest itself differently in rural and urban
environments in Nepal is in terms of seasonal variation. However, like the results of
seasonal growth variation among rural Nepalese children found by Costello (1989)
and Panter-Brick (1997), the most pronounced effect of seasonality was the
significantly lower mean WHZ scores during the summer months. Costello (1989)
attributes this drop in weight to the dearth in food availability before the harvest;
Panter-Brick (1997) examines a combination of factors including harvest cycles and
child illness. Given that the peri-urban children in this sample are completely
dependent on the urban food markets, i.e. there is no pre-harvestlpost-harvest
difference in food availability, it is most likely that seasonal differences are not due
to changes in food consumption. Rather, it is probably associated with the increased
incidence and severity of illness, particularly gastrointestinal infections, during the
hot and monsoon seasons. The peak in diarrhoeal diseases during these seasons is
documented in other low-income countries located in tropical regions of the world
(Rowland and Barrell 1980, Rowland 1983).
GROWTH STUNTING IN NEPALESE CHllDREN 265
5. CONCLUSION
This paper examined linear growth retardation or stunting among a particular group
of children residing in peri-urban Kathmandu, Nepal. The two main goals were: 1) to
examine linear growth faltering in relation to other types of body growth; and 2) to
evaluate whether growth among children in peri-urban Nepal is uniquely different
from that among children in both urban and rural Nepal.
It appears that the process of stunting is very different from being stunted.
Decreased growth velocity occurs from infancy into the second year of life; however,
the children do not appear to be stunted (as defined by the -2 SD cut-off) until they
are over 24 months of age. This underscores the importance of examining factors
that affect stunting - such as macro and micro-nutrient deficiencies as well as
infectious disease - early on in infancy, before the signs of stunting, in terms of
cross-sectional growth measures, actually surface. Other anthropometrics such as
weight and arm measurements shed light on linear growth faltering. Although for the
most part among the sample wasting is not an issue, there is a marked increase in
low mean WHZ among 12 to 24 month old children, suggesting that acute weight
loss may precede or accompany linear growth retardation. Low mean upper arm
muscle mass (ZAM) mirrors the state of being stunted in this sample, indicating
chronic nutrient deprivation. Thus it is important to examine more than just arm
circumference to get a sense of growth deficits among children experiencing growth
stunting.
Comparisons between this peri-urban sample of children and other rural and urban
samples in Nepal show that the prevalence of stunting, although very high among
the peri-urban children, is more in line with urban than rural children. Both urban
and peri-urban children from six to 36 months of age have a significantly lower
prevalence of stunting than rural children. Despite problems of peri-urban
settlements such as lack of sanitation and poverty, there are still some benefits
accrued to children in terms of growth and development when living in an urban
environment compared to a rural village. Interestingly, despite differences in food
procurement between rural and peri-urban people, there are still seasonal differences
in growth among this peri-urban sample of children. Like in rural populations, mean
WHZ was lower in the hot and monsoon seasons compared to the cold season.
Seasonal differences in weight gain and loss are probably more related to infectious
disease than to seasonal differences in food consumption.
Finally, despite the fact that this sample of children is a unique sub-population
living in a particular peri-urban location in Nepal, it is striking how similar the
phenomenon of stunting is among children living in different locations around the
world. The common denominator among all of these groups is poverty.
6. REFERENCES
Allen, L.H., 1994, Nutritional influences on linear growth: a general review. European Journal of
Clinical Nutrition, 48 (Supp!. I), S75-S89.
Anderson, M., 1979, Comparison of anthropometric measures of nutritional status in preschool children
in five developing countries. American Journal of Clinical Nutrition, 32, 2339-2345.
Baumgartner, R.N., Roche, A.F. and Himes, J.H., 1986, Incremental growth tables: supplementary to
previously published charts. The American Journal of Clinical Nutrition, 43, 711-722.
266 CHAPTER 20
Beaton, G.H., 1992, The Raymond Pearl Memorial Lecture, 1990: Nutrition research in human biology:
changing perspectives and interpretations. American Journal of Human Biology, 4, 159-177.
Brink, E.W., Khan, I.H., Splitter, J.L., Staehling, N.W., Lane, J.M. and Nichaman, M.Z, 1976,
Nutritional status of children in Nepal, 1975. Bulletin of the World Health Organization, 54, 311-
318.
Brockerhoff, M., 1995, Child survival in big cities: the disadvantages of migrants. Social Science and
Medicine, 40, 1371-1383.
Chen, L.C., Chowdhury, A. and Huffman, S.L., 1980, Anthropometric assessment of energy-protein
malnutrition and subsequent risk of mortality among preschool aged children. The American
Journal of Clinical Nutrition, 33,1836-1845.
Costello, A.M., 1989, Growth velocity and stunting in rural Nepal. Archives of Disease in Childhood,
64, 1478-1482.
Dettwyler, K.A., 1991, Growth status of children in rural Mali: implications for nutrition education
programs. American Journal of Human Biology, 3, 447-462.
Eveleth, P.B. and Tanner, J.M., 1990,Worldwide Variation in Human Growth (2nd ed.) (Cambridge:
Cambridge University Press).
Gittelsohn, J., Thapa, M. and Landman, L.T., 1997, Cultural factors, caloric intake and micronutrient
sufficiency in rural Nepali households. Social Science and Medicine, 44,1739-1749.
Harpham, T., Lusty, T. and Vaughan, P., 1988, In the Shadow of the City. Community Health and the
Urban Poor (Oxford: Oxford University Press).
Himrnelgreen, D.A., Dannenhoffer, R., Baht, I. and Lee, R.V., 1991, Anthropometric assessment of
nutritional status among highland Kashmiri children: reevaluating the assumption of female
nutritional disadvantage. American Journal of Human Biology, 3, 239-249.
Frisancho, A.R., 1990, Anthropometric Standards for the Assessment of Growth and Nutritional Status
(Ann Arbor: University of Michigan Press).
Gorstein, J., Sullivan, K., Yip, R., de Onis, M., Trowbridge, F., Fajans, P. and Clugston, G., 1994, Issues
in the assessment of nutritional status using anthropometry. Bulletin of the World Health
Organization, 72, 273-283.
Keller, W., 1988, The epidemiology of stunting In Linear Growth Retardation in Less Developed
Countries. Nestle Nutrition Workshop Series Volume 14, edited by J.C. Waterlow (New York:
Raven Press) pp. 17-40.
Leonard, W.R., 1991, Household-level strategies for protecting children from seasonal food scarcity.
Social Science and Medicine, 33, 1127-1133.
Leonard, W.R., DeWalt, K.M., Stansbury, J. P. and McCaston, M.K., 1995, Growth differences
between children of highland and coastal Ecuador. American Journal of Physical Anthropology,
98,47-57.
Lohman, T.G., Roche, A.F.and Martorell, R., 1988, Anthropometric Standardization Reference Manual
(Campaign, Illinois: Human Kinetics Books).
Martorell, R., Khan, L.K. and Schroeder, D.G., 1994, Reversibility of stunting: epidemiological findings
in children from developing countries. European Journal of Clinical Nutrition, 48 (Suppl. 1), S45-
S57.
Martorell, R., Leslie, J. and Moock, P.R., 1987, Characteristics and determinants of child nutritional
status in Nepal. The American Journal of Clinical Nutrition, 39, 74-86.
Naborro, D., Howard, P., Cassels, e., Pant, M, Wijga, A. and Padfield, N., 1988, The importance of
infections and environmental factors as possible determinants of growth retardation in children. In
Linear Growth Retardation in Less Developed Countries. Nestle Nutrition Workshop Series
Volume 14, edited by J.C. Waterlow (New York: Raven Press) pp. 165-184.
Neumann, e.G. and Harrison, G.G., 1994, Onset and evolution of stunting in infants and children.
Examples form the Human Nutrition Collaborative Research Support Program. Kenya and Egypt
studies. European Journal of Clinical Nutrition, 48 (Suppl. 1), S90-S102.
Panter-Brick, C., 1997, Seasonal growth patterns in rural Nepali children. Annals of Human Biology,
24,1-18.
Panter-Brick, C., Todd, A. and Baker, R., 1996, Growth status of homeless Nepali boys: Do they differ
from rural and urban controls? Social Science and Medicine, 43, 441-451.
Ricci, J.A. and Becker, S., 1996, Risk factors for wasting and stunting among children in Metro Cebu,
Philippines. American Journal of Clinical Nutrition, 63, 966-975.
Rowland, M.G.M., 1983, Epidemiology of childhood diarrhea in the Gambia. In Diarrhea and
Malnutrition edited by L.C. Chen and N.C. Scrimshaw (New York: Plenum Press) pp. 87-98.
Rowland, M.G.M. and Barrell, R.A.E., 1980, Ecological factors in gastroenteritis. In Disease and
Urbanization, edited by EJ. Clegg and J. P. Garlic (London: Taylor and Francis Ltd) pp. 21-35.
Sastry, J. G., Vijayaraghavan, K. and Rao, N. P., 1989, Indian preschool children - a profile of stunting
and wasting. Journal of Tropical Pediatrics, 35, 237-240.
UNICEF, 1992, Children and Women of Nepal. A Situation Analysis (Kathmandu: UNICEF - Nepal).
UNICEF, 1996, Nepal Multiple Indicator Surveillance Health and Nutrition - Cycle 1, January to March
1995. (Kathmandu: UNICEF - Nepal).
GROWTH STUNTING IN NEPALESE CHll.DREN 267
Victora, CG., 1992, The association between wasting and stunting: an international perspective. Journal
of Nutrition, 122, 1105-1110.
Waterlow, J.C., 1988,Observations on the natural history of stunting. In Linear Growth Retardation in
Less Developed Countries. Nestle Nutrition Workshop Series Volume 14, edited by J.C. Waterlow
(New York: Raven Press) pp. 1-16.
Waterlow, J.C., 1992, Protein Energy Malnutrition (London: Edward Arnold).
Waterlow, J.C, 1994,Causes and mechanisms of linear growth retardation (stunting). European Journal
of Clinical Nutrition, 48 (Suppl. I), S I-S4.
Watts, T., Ng'andu, N. and Wray, J., 1990, Children in an Urban Township in Zambia. A prospective
study of children during their first year of life. Journal of Tropical Pediatrics, 36, 287-293.
WHO, 1988, Urbanization and its Implications for Child Health (Geneva: World Health Organization in
collaboration with the United Nations Environment Program).
Williams, B.T., 1990,Assessing the health impact of urbanization. World Health Statistics Quarterly, 43,
145-150.
CHAPTER 21
1. INTRODUCTION
Populations of the South Seas are broadly classified as Polynesian, Melanesian and
Micronesian (Campbell 1989). However, their origin is relatively homogeneous in
an archaeological sense (Gravelle 1979). The distance between Fiji (Melanesia) and
Samoa (Polynesia) is not great. However, the populations have been subjected to
very different processes of socio-cultural change in the past generation. This, in turn,
has affected the growth status of the children. Complex interactions between genetic
and environmental factors contribute to variation between populations. Comparisons
among South Seas populations provide an opportunity to consider this variation
especially in children.
The growth status of children in the South Seas area has been documented,
including studies in Fiji (Hawley and Jansen 1971, Lourie 1972, Clegg 1989) and
Samoa (Wigg 1978, Baker et al. 1986, Bindon and Zansky 1986). However,
anthropometric data on these ethnic groups are still quite limited.
This paper focuses on the Skelic Index, which assesses the relationship of lower
limb length to sitting height. The Skelic Index changes during prepubertal growth
(greater contribution of lower limb length) and during pubertal growth (greater
contribution of sitting height) (Cameron et al. 1982, Malina and Bouchard 1991).
This study contrasts the pattern of growth in the Skelic Index among Fijian and
Samoan children during childhood and adolescence, and examines ethnic and sex
differences in body size and the index in cross-sectional sample of Fijian school
children resident in an urban area and Samoan school children resident in a rural
area.
girls) and 192 Samoan school children (95 boys and 97 girls) 6+ to 17+ years of
age. Chronological age (decimal age) was computed from birth dates and the date of
observation. Stature was measured in the standard erect posture on barefoot subjects,
and sitting height was measured in the standard sitting posture on a table top. A
field anthropometer was used and measurements were recorded to the nearest
millimetre. All measurements were done by the same individual (TS). Lower limb
length was calculated by subtracting sitting height from stature, and the Skelic Index
was calculated as lower limb length/sitting height.
Comparisons of within ethnic groups and between sexes were carried out using
analysis of covariance with age as the covariate. Analyses focused on two specific
comparisons: first, comparisons between and within Fijian and Samoan children in
each sex: second, comparisons between boys and girls in each ethnic group. All
statistics were calculated using the SAS statistical package.
3. RESULTS
Tables are shown in the Appendix. Descriptive statistics for Fijian and Samoan
children are given in Tables 1 and 2, respectively. There are few significant (p<0.05)
mean differences between boys and girls in both Fijian and Samoan children (Tables
1 and 2). There is only one significant difference in the Skelic Index in Samoan (at
7+ years). On the other hand, Fijian boys and girls significantly differ at 7+, 8+,
12+, 14+, 15+, 16+, and 17+ years of age.
Mean differences between Fijian and Samoan children within each sex (Table 1)
are small and very few are significant in stature, sitting height and lower limb
length. However, ethnic differences within sex in the Skelic Index are often
significant. The Skelic Indices of Samoan boys are significantly greater than those of
Fijian boys at 6+, 7+, 8+, 9+ and 13+ years of age. The same trend is evident in
girls, 6+, 7+, 8+, 9+, 13+ and 16+ years of age.
Figures 1 to 4 are scatter plots of stature, sitting height, and the Skelic Index by
chronological age. As expected, stature and sitting height increase with age (Figures
1 and 2). The Skelic Index shows ethnic variation by ethnicity and sex, respectively
(Figures 3 and 4).
Results of the analyses of covariance, with age as the covariate, are summarised
in Table 3. There are no significant interactions between ethnicity and sex in the
anthropometric dimensions and the Skelic Index. There are also no significant ethnic
differences in stature and lower limb length. Sitting height and the Skelic Index, in
contrast, show significant ethnic differences. Sex differences are significant for all
variables, except the sitting height.
Directions of the differences for Fijian and Samoan children are shown in Table
4, which gives age-adjusted means. Fijian children have a significantly longer
sitting height than Samoan children, whereas Samoan children have significantly
greater Skelic Index. And comparisons of the age-adjusted means between boys and
girls show that boys are significantly taller and have a longer lower limb length and
greater Skelic Index than girls. There are no significant sex differences in sitting
height.
SKELIC INDEX IN F IJIAN AND S AMOAN CHILDREN 271
200
~
oa
150 f-
.. . 1 .:
if~*~iof. 0
•
0
,.~ t4
•.
0
Stature
6
(,)
I
Q)~ 100
N
U5 t ~--j
50 f-
MIJI" ~ Sitting height
o Boys
• Girls
o t
4 6
I I
8
I
10
I
12
I
14
I
16 18
I
20
Age, years
Figure 1. Age changes and sex differences in stature and sitting height in Fijian school
children
200
~
150 f-
0.. ~ o ~~
o ~ ~ •• ~"oooot • , •
•
.1. 0 •
0
au
,. • r- • 0 ~ Stature
~~
..... 100 ~
V)
.. ~••~ i""'~~
S· .
.~ro
Ithng height
50 f- o Boys
• Girls
o t
4 6
I I
8
I
10
I
12
I
14
I
16 18
I
20
Age, years
Figure 2. Age changes and sex differences in stature and sitting height in Samoan
school children
272 CHAPTER 21
110
.g
.S
~
u
:..::l
100
90
~~,
o ~~ ~o~
~.
..
••
~
~_
tIP.
••• 0. Z' J
J .
~ 80 8 0
U)
Fijian children
70
120
o Boys
~ 110
t ..,.o.~t9.~~~~ i9.~oo~8e ~.
11)
• Girls 0 0
]
u
100 • 0 0 ~;.~
:..::l 90 •• lj 0 " • • 0
~
U)
• • o. • \
80 Samoan children
70
4 6 8 10 12 14 16 18 20
Age, years
Figure 3. Age changes and sex differences in the Skelic Index in Fijian and Samoan
school children
110
° eO •
.g • JI'~~' o ~. ~ ~
~
100 .' .,.1....
.S
u
:..::l
90
t80':·~i~ I~· ~~~.~.o.JI, fiJ
~ 80 8 0 o
° 00
U)
Boys
70
120
o Fijian
110
•• •.01,
~
.g • Samoan
.S 100 • c\t.. it~ ..• ~
~~ co'·~~ ~ O ~g"~ _D
u · 0
:..::l 90
~
U) ou
- lj 0 0 ~ 0Q:l 0 ~ ~ ~
80 Girls 0
70 I~~_ _~_ _~_ _~~_ _~_ _~_ _~~_ _- L_ _~_ _~~_ _~_ _~~
4 6 8 10 12 14 16 18 20
Age, years
Figure 4. Age changes and ethnic differences in the Skelic Index in boys and girls
SKELIC INDEX IN FIJIAN AND SAMOAN CHILDREN 273
4. DISCUSSION
Although numbers of children in each age group are limited, the data suggest several
trends in the comparison of growth in the Fijian and Samoan populations. Fijian
and Samoan children have similar statures, but Fijian children, especially in girls,
tend to exceed Samoan children in sitting height and Samoan children tend to
exceed Fijian children in lower limb length (Tables 1 and 2). The Fijian sample is
from an urban area and Samoan sample is from a rural area. It is reasonably well
documented that an urban children are larger in body size than the children living in
a rural area (Dasgupta and Das 1997, Bogin 1999). Compared with both the same
ethnic samples of different areas from the present study, results may differ from this
study.
The data of Samoan children are from rural villages on the island of Savaii.
Bindon and Zansky (1986) reported on growth changes of stature and body weight
on Western Samoa for similar rural villages, while Clegg (1989) reported several
physical characteristics of children in Suva, Fiji. Compared to these earlier studies,
statures in the present sample are, on average, greater in children of both ethnic
groups at almost ages. The secular increase in stature has been widely reported in
many different ethnic groups (Eveleth and Tanner 1990, Ulijaszek 1993). This
results thus suggest a similar secular change in stature in both ethnic groups.
The Skelic Index is, on average, greater in Samoan than in Fijian children in
each sex (Tables 1 and 2). Growth changes of the Index are, however, different in
both ethnic groups. Differences in relative leg length are evident among other ethnic
groups (Krogman 1970, Malina et al. 1974, Meredith and Spurgeon 1976, Meredith
1979, Buschang et al. 1986, Eveleth and Tanner 1990, Nath et al. 1991, Satake et
al. 1995, Spurgeon et al. 1997).
The Skelic Indices of boys are greater than those of girls in each ethnic group.
Absolute differences in the Skelic Index of boys in Fijian children increase in the
prepubertal period. And after that Fijian children, but not Samoan children, keep
differences between both sexes (Tables 1 and 2, and Figure 3). The sex difference in
the Skelic Index is significant (Tables 3 and 4) and the sex difference in age-adjusted
means is greater in Fijian than in Samoan children (Table 6). The results suggest
that during adolescence, Samoan children are proportionately similar, while Fijian
girls have relatively shorter lower limb length than boys. In both ethnic groups,
however, girls have a proportionately shorter lower limb length than boys as
Meredith and Knott (1938) and Buckler (1990) have also shown. Satake, Oh, Boo
and Hattori (1995) also showed sexual dimorphism in age changes of the Skelic
Index among American, Japanese and Korean children. The relationship between
boys and girls in age change of the Skelic Index differed within each ethnic group.
Lower limb length experiences its adolescent growth spurt before sitting height
(Tanner et al. 1982). And, the timing of prepubertal and pubertal growth in
anthropometric dimensions varies among different ethnic groups and between both
sexes (Malina et al. 1988). As a result, it is anticipated that age change in the Skelic
274 CHAPTER 21
Index would show different trends among ethnic groups and between sexes. Sexual
dimorphism in age changes of the Skelic Index may be useful in understanding
growth differences among ethnic groups.
The lack of significant interactions between ethnicity and sex in any of the
anthropometric dimensions or the Skelic index is of interest (Table 3). It suggests
that the ethnic influence on the anthropometric characteristics and the index is
independent of sex differences, and also that sex differences influence individuals from
both ethnic groups in the same way.
Ethnic comparisons within sex show that Fijian children were a significantly
longer sitting height than Samoan children, and that Samoan children have a
significantly greater Skelic Index (Table 3 and 4). The presence of significant ethnic
differences in sitting height and the lack of significant differences in stature and lower
limb length contribute to the significant difference in the Skelic Index.
Regardless of ethnicity, sex differences in stature, lower limb length and the
Skelic Index are significant. Sitting height, on the other hand, dose not differ
between boys and girls (Tables 3 and 4). The lack of significant sex differences in
sitting height and the presence of significant differences in stature and lower limb
length contribute to the significant sex difference in the Skelic Index. As such,
growth of sitting height is an important variable both in ethnic differences between
Fijian and Samoan children and in sexual dimorphism of linear body proportions.
Acknowledgement. The authors express their sincere gratitude to Dr. R.M. Malina
of Michigan State University for variable comments and helpful advice in preparation
of the English manuscript. The present study was financially supported by Nihon
University.
5. REFERENCES
Baker, P.T., Hanna, J.M., and Baker, T.S., 1986, The changing Samoan, (New York: Oxford University
Press).
Bindon, J.R., and Zansky, S.M., 1986, Growth patterns of height and weight among three groups of
Samoan preadolescents. Annals of Human Biology, 13, 171-178.
Bogin, B., 1999, Patterns of Human Growth (Cambridge: Cambridge University Press)
Buckler, J.M., 1990, A longitudinal study of adolescent growth. (London: Springer Verlag).
Buschang, P.H., Malina, R.M., and Little, B.B., 1986, Linear growth ofZapotec school children: growth
status and yearly velocity for leg length and sitting height. Annals of Human Biology, 13,225-234.
Cameron, N., Tanner, J.M., and Whitehouse, R.H., 1982, A longitudinal analysis of the growth of limb
segments in adolescence. Annals of Human Biology, 9, 211-220.
Campbell I.C., 1989, A History of the Pacific Island. (New Zealand: Canterbury University Press).
Clegg, EJ., 1989, The growth of Melanesian and Indian children in Fiji. Annals of Human Biology, 16,
507-528.
Dasgupta, P., and Das, S.R., 1997, A cross-sectional growth of trunk and limb segments of the Bengali
boys of Calcutta. Annals of Human Biology, 24, 363-369.
Eveleth, P.B., and Tanner, J.M., 1990, Worldwide variation in human growth. (Cambridge: Cambridge
University Press).
Gravelle, K., 1979, Fiji's Times A History of Fiji. (Fiji: The Fiji Times Limited)
Hawley, T.G. and Jansen, A.AJ., 1971, Height and weight of Fijians in coastal areas from one year to
adulthood. New Zealand Medical Journal, 73, 346-349.
Krogman, W.M., 1970, Growth of head, face trunk, and limbs in Philadelphia white and negro children
of elementary and high school age. Monographs of the Society of Research in Child Development,
35: serial No. 136, 1-80.
SKELIC INDEX IN FIJIAN AND SAMOAN CHILDREN 275
Lourie, lA., 1972, Anthropometry of Lau Islanders, Fiji with a note on their color vision. Human
Biology in Oceania, 1, 273-277.
Malina, RM., and Bouchard, C., 1991, Growth, Maturation, and Physical Activity. (Champaign, IL:
Human Kinetics).
Malina, R.M., Bouchard, c., Beunen, G., 1988, Human growth: Selected aspects of current research on
well-nourished children. Annual Reviews Anthropology, 17, 187-219.
Malina, RM., Hamil, P.V.V., and Lemeshow, S., 1974, Body dimensions and proportions, White and
Negro children 6-11 years, United States. Vital and Health Statistics, Series 11, Number 143, 66p.
Washington, D.C.: National Center for Health Statistics.
Meredith, H.V., 1979, Relationships oflower limb height to sitting height in black populations of Africa
and United States. American Journal of Physical Anthropology, 51, 63-66.
Meredith, H.V., and Knott, V.B., 1938, Changes in body proportions during infancy and the preschool
years III. The Skelic index. Child Development, 9, 49-62.
Meredith, H.V., and Spurgeon, lH., 1976, Comparative findings on the skelic index of black and white
children and youths residing in south Carolina. Growth, 40, 75-81.
Nath. S., French, K.E., and Spurgeon, J.H., 1991, Somatic comparisons: Baiga and Gond males of
Madhya Pradesh, India. American Journal of Human Biology, 3, 281-287.
Satake T., Yoshida K., Kanazawa E., Hattori K., 1996, Ethnic difference in the Skelic Index among
school children of Western Samoa and Fiji. Auxology 3, 28-30 (in Japanese).
Satake, T., 1996, Anthropological Study on Food intake and Oral Morphology of Polynesians in Special
Oceanic Environment. Final Report of the Interdisciplinary General Joint Research Grant for
Nihon University 1994-1996, pp. 106-118.
Satake, T., Oh, S.I., Boo, K.W., and Hattori, K., 1995, Growth of Korean schoolchildren using Skelic
Index. Bulletin of College of General Education Ibaraki University, 28,257-268(in Japanese with
English summary).
Spurgeon, lH., Kim, K-B., French, K.E., Giese, W.K., 1997, Somatic comparisons at three ages of
south Korean males and males of other Asian groups. American Journal of Human Biology, 9, 493-
503.
Tanner, J.M., Hayashi, T., Preece, M.A., Cameron, N., 1982, Increase in length ofieg relative to trunk
in Japanese children. Annals of Human Biology, 9, 411-423.
Ulijaszek, S.J., 1993, Evidence for a secular trend in heights and weights of adults in Papua New
Guinea. Annals of Human Biology, 20, 349-355.
Wigg, N .R, 1978, The anthropometry of Western Samoan pre-school children. New Zealand Medical
Journal, 87, 172-175.
276 CHAPTER 21
APPENDIX
Table 1. Descriptive statistics for stature, sitting height, lower limb length and the
Skelic Index by age and sex in Fijian school children (significances of t-Student test)
Girls (N = 1232
Stature, cm Sitting height, Lower Limb Skelic Index
em length, cm
Age, n Mean SD Mean SD Mean SD Mean SD
years
6+ 8 119.4 2.91 64.2 1.43# 55.2 1.95 86.0 2.75#
7+ 11 125.6 3.63 66.9 1.97 58.7 2.03 87.7 2.40##
8+ 9 129.8 4.11 68.2 2.25## 61.6 1.96 90.3 1.33##
9+ 10 136.1 5.58 71.0 3.32 65.0 2.65 91.6 2.94##
10+ 2 145.1 0.35 75.5 1.70 69.6 1.34 92.2 3.85
11+ 12 146.1 5.37 76.0 3.75 70.1 2.46 92.5 4.00
12+ 16 156.3 5.97 81.9 3.71 74.5 3.22 91.1 4.15
13+ 13 160.6 8.70 83.6 5.21 77.0 4.48 92.2 4.77#
14+ 17 158.4 5.19# 82.3 3.66 76.0 2.31 92.4 3.66
15+ 11 161.6 3.92 84.8 1.87 76.8 2.89 90.5 3.34
16+ 6 163.2 8.70 86.1 3.35 77.2 5.59 89.6 3.74#
17+ 8 166.0 5.13 87.2 1.78 78.8 3.95 90.3 3.88
Table 2. Descriptive statistics for stature, sitting height, lower limb length and the
Skelic Index by age and sex in Samoan school children (significances of t-Student test)
Girls (N = 972
Stature, cm Sitting height, Lower Limb Skelic Index
cm length, cm
Age, n Mean SD Mean SD Mean SD Mean SD
years
6+ 9 117.0 4.10 61.5 2.29 55.5 2.51 90.3 4.07
7+ 9 125.0 6.52 65.2 2.97 59.8 3.60 91.6 1.92
8+ 5 128.0 2.10 64.4 0.86 63.6 2.56 98.6 2.56
9+ 11 136.9 6.76 70.0 3.46 66.9 3.60 95.6 2.86
10+ 6 143.7 4.21 73.3 2.72 70.4 2.03 96.1 3.04
11+ 4 145.1 l3.42 73.5 6.41 71.6 7.31 97.3 4.23
12+ 3 156.3 9.49 80.4 5.73 75.9 4.03 94.5 3.11
13+ 9 157.2 5.80 80.1 2.98 77.1 3.50 96.3 3.59
14+ 5 163.7 4.07 85.3 3.20 78.4 3.14 92.0 3.14
15+ 13 162.4 5.15 83.9 2.52 78.5 3.87 93.7 4.68
16+ 19 165.6 4.49 85.1 2.61 80.5 2.91 94.6 3.70
17+ 4 164.0 7.06 85.9 3.23 78.1 5.46 90.1 6.42
* P < 0.05
** P < 0.01
278 CHAPTER 21
Table 3. Results of analysis of covariance, with age as the covariate, of stature, sitting
height, lower limb length and the Skelic Index of Fijian and Samoan school children
MS F p
Stature Ethnie(E)a 108.07 2.52 NS
Sex(S)a 682.55 15.92 <0.01
E*Sa 3.68 0.09 NS
Residual(R)b 42.86
Sitting height Ethnie(E)a 263.89 22.39 <0.01
Sex(S)a 12.05 1.02 NS
E*Sa 17.45 1.48 NS
Residual(R)b 11.79
Lower limb Ethnie(E)a 34.21 2.22 NS
length Sex(S)a 513.24 33.32 <0.01
E*Sa 5.10 0.33 NS
Residual(R)b 15.40
Table 4. Age-adjusted means of stature, sitting height, lower limb length and the Skelic
Index of Fijian and Samoan school children
Ethnicity Sex
Fijian Samoan Boys Girls
Stature, em 149.6 148.6 150.4 147.8
Sitting height, em 77.8 76.2 77.2 76.8
Lower limb length, em 71.8 72.3 73.1 71.0
Skelie Index 92.1 94.9 94.6 92.4
SKELIC INDEX IN FIJIAN AND SAMOAN CHILDREN 279
Fiji Samoa
MS F P MS F P
Stature Sex(S)a 339.09 7.00 <0.01 346.87 9.70 <0.01
Residual(R)b 48.56 35.77
Sitting height Sex(S)a 0.05 0.00 NS 31.20 3.16 NS
Residual(R)b 13.16 9.88
Lower limb length Sex(S)a 347.74 20.42 <0.01 170.02 12.94 <0.01
Residual(R)b 17.03 13.14
Skelie Index Sex(S)a 519.69 28.31 <0.01 103.23 6.27 <0.01
Residual~R2b 18.36 16.47
adf= 1
bdf= 248 Fijian, 189 Samoan
Table 6. Age-adjusted means of stature, sitting height, lower limb length and the Skelic
Index of Fijian and Samoan boys and girls
Fijian Samoan
Boys Girls Boys Girls
Stature, em 150.5 148.2 150.3 147.6
Sitting height, em 77.7 77.7 76.9 76.0
Lower limb length, em 72.8 70.5 73.5 71.6
Skelie Index 93.6 90.7 95.6 94.2
CHAPTER 22
K. KIMURA
Kimura Auxological Institute, Sayama, Japan
1. INTRODUCTION
The present study forms a part of comprehensive anthropological investigation on
children of mixed American and Japanese parentage, carried out from 1949 to 1965
longitudinally. They were all the offspring of American Servicemen of the
Occupation Forces and Japanese women, born after the World War II. They were
cared for at the Elizabeth Sanders Home in Ohiso and the Boy's Town in Yamato,
Kanagawa. For the study, according to the father's racial characteristics, they were
divided into two groups of mixed Afro- and Euro-American and Japanese parentage.
It was rather difficult to obtain accurate information about the father's racial
background. They were placed into each group mainly on the basis of the mother's
statement, and also on their outstanding physical traits, when obvious.
The late Professor Suda has stated a general introduction (Suda 1968). Some co-
workers charged of the survey of their deciduous teeth (Hanihara 1954-57, 1963,
1965, 1968), oral zone (Hojo 1964), hair, iris and skin colors (Goto 1961, Honda
1958, Michibe 1961, Omoto 1968), facial muscles (Kondo et al. 1968) and finger
and palm prints (Kimura 1974a, 1974b). The others examined the urinary BAIB
excretion (Ishimoto 1968), blood groups (Matsuzawa and Kitamura 1956, Suzuki
and Matsuzawa 1956, Kitamura and Suzuki 1956), and subcutaneous fat (Kohara
1968). Physical measurements were measured longitudinally (Hoshi 1959, 1970,
Suda et at. 1965, 1968, 1973, 1975, 1976a, 1976b). I had taken radiographs ofthe
hand and wrist since 1958, and reported preliminarily on the skeletal maturity of the
children of only Euro-American and Japanese parentage, using the Oxford method
(Acheson 1954) and the TWI and TW2 method (Tanner et al. 1962, 1983, Kimura
1971, 1972, 1976a, 1994). I have also studied on the non-epiphyseal notch (Kimura
et at. 1982) and growth of the second metacarpal of both groups of mixed American
and Japanese parentage (Kimura 1996). The present paper examines skeletal maturity
of the children of mixed Afro- and Euro-American and Japanese parentage, compared
with the Japanese, using the TW2 method (Tanner et al. 1983).
281
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 281-297.
© 200t Kluwer Academic Publishers.
282 CHAPTER 22
3. RESULTS
Figure 1 shows the pattern of differences between the means of skeletal age (20-bone,
RUS and carpal) and chronological ages (SA-CA) plotted against the means of each
chronological age in the children of mixed Afro- and Euro-American and Japanese
parentage and the war orphans, including the Japanese control. The mid-age method
was used fro chronological age grouping: for example, the 7 years group contains the
children aged 6.5 to 7.4 years at the date of investigation.
Common to each skeletal age, the control was somewhat delayed in maturity
before 12 years in boys and 9 years, respectively, and then rather advanced in both
sexes, compared with the British standards. The war orphan became gradually
delayed in maturity after 10 years of age in the boys, especially in the carpal age, but
followed roughly a similar pace of maturity in the girls, compared with the control.
The children of mixed Afro-American and Japanese parentage were slightly delayed
until 9 years of age and then advanced in the boys, and always advanced after 6 years
of age, markedly at 10 years, in the girls, with the British standards. The children of
mixed Euro-American and Japanese parentage were delayed until 10 years of age and
SKELETAL MATURITY IN CHILDREN OF MIXED PARENTAGE 283
then markedly advanced in the boys, and comparable at 6-9 years of age and then
markedly advanced in the girls, with the standards. They all arrived at full maturity
l
for each skeletal age at the same age for each TW2 age, as the British standards.
BOYS
§2 rO- bone
age ~!~
}t
,0 ~ _-o "
~ -~.,
~- .. ..
~~~
'7o~~
l RUS age
~ G- - ,,-' ". .•.• • 0 •
-2
2 , Carpal age
«
C? 0 I ~ ......-- __ .~o ~ ' A' " ' i '~"'"
~
-2 4 5 6 l' 8 9 1 '0 1 '1 1'21'3141'5 1'6 171819
l
Chronological age, years
GIRLS
§2r~
..
«W,0 0- -
'6--' .
" .,
l
o'
••
§-~t ~
~
'
.. 0.'
-2
, , -- .. '
q.". • .
«
Ir
-2 ' p
4
Figure 1. Age changes of the difference between TW2 skeletal and chronological ages
in the children of mixed American and Japanese parentage and the Japanese war-
orphans and controls
284 CHAPTER 22
4. DISCUSSION
Most variations in skeletal maturation are genetically determined (Pryor 1907, 1936,
1939, Hess and Abramson 1933, Buschke 1935, Key 1936, Rigler 1938, Reynolds
1943, Gam et al. 1961a, 1961b, 1962a, 1962b, 1963, 1969, Hertzog et al. 1969,
Kimura 1962, 1981, 1983, Sklad 1973). However population differences in physique
and growth may reflect not only genetic factors, but also environmental ones.
assessed by the author alone on each radiograph of all samples twice and more at
intervals of a few weeks and more. In the longitudinal samples, each individual
series was assessed successively.
1984), except for those in developing and poor countries (Meredith 1984, Prado et
al. 1986, Preece 1989, Rosique and Rebato 1995).
Comparison of the data of the British by Poland (1898) and Stuart et al. (1962)
indicated a secular increase in the rates of skeletal maturation of approximately 0.22-
0.66 year/decade (Himes 1984). But no changes were found in skeletal maturity
corresponding to the secular trend in the United States (Roche and Davila 1976,
Roche et al. 1978). In Japan, it seems that the secular trend to toward greater stature
since 1900 arrived at a final stage approximately after 1970 in students and 1980 in
the general public, and that the secular trend proceeded rapidly in a population under
conditions of vigorous socio-economic changes (Kimura 1984). Few differences have
been already evident in the height and body weight of children between the hilly and
the lowland sections of Tokyo in 1995 (Kimura 1999). However, an acceleration of
the RUS maturation was observed in Japanese children between 1968-69 and 1986
(Ashizawa 1994).
West African boys were delayed by an average 1.5 years after 9 years of age (Weiner
and Thambipillai 1952), and South Africans in Pretoria by an average 0.7 years in
boys and 1.5 years in girls in 7-12 years of age (Levine 1972), with the Greulich and
Pyle standards. According to a longitudinal study in West Africans (TWl), the girls
and boys both were retarded at 11 years of age but caught up during puberty. The
girls actually reached and the boys were only 0.5 years retarded the standards at 15
years of age (Michaut et al. 1972). African Coloured children showed skeletal
advancement in the early years of life over Europeans (Dommisse and Leipoldt
1936). Pretoria Black and Coloured children did not differ significantly from each
other (Levine 1972). The mean TW2 ages are always less in children ofKhoi Khoi
and Rehoboth Basters than the British standards by one or two years in both sexes
(Singer and Kimura 1981). Thus, in general, African children were advanced
skeletally in infancy (Hess and Weinstock 1925, Beresowski and Lundie 1952,
Falkner 1958), followed by deceleration and finally retardation in middle and late
childhood (Masse and Hunt 1963), compared with European children. Africans
apparently have a different maturational pattern from Europeans as far as timing and
duration of some stages is concerned (Wingerd et al. 1974). The rates of skeletal
maturation for African children in the United States (Roche et al. 1975) are more
rapid than in Africa (Levine 1972) and also in Jamaica (Marshall et al. 1970).
European descendants in Pretoria (Levine 1972) and Melbourne (Roche 1967)
were closely comparable with the Greulich-Pyle standards. The Euro-American
children indicated a trend to lesser retardation in skeletal maturity with increasing
age in Atlanta (Gray and Lamons 1959), Philadelphia (Johnston 1962) and Nebraska
(Fry 1966). The same trend was observed in the French Canadians (Baughan et al.
1979). The National Health Survey (NHS) in the United States in 1963-70 found
that the Euro-American boys and girls are retarded between 9 and 13 years of age and
that the definite tendency increases retardation (Roche et al. 1974, 1976). In
Philadelphia, however, the boys were first skeletally retarded, but soon more ahead
of the Greulich-Pyle standards to a peak advancement at 11 years of age. But the
girls did not move ahead in the standards until more later, beginning this spurt at
10 years, and levelling off at 14 years of age (Johnston 1963). The Euro-Americans
tended to be slightly advanced in skeletal maturity status with the British standards
(Malina 1970). The British descendants in Australia also accelerated at about 11
years in the boys and 9 years in the girls, after 12 years of age they decelerated in the
girls (Roche 1967), relative to the British standards.
First, Todd (1931) noted that the Afro-American girls are often in advance
skeletally to the Euro-American girls. A similar precocity in ossification of the
skeleton was demonstrated for the African and Afro-American new-borns and infants
compared with the Euro-Americans (Kelly and Reynolds 1947, Christie 1949,
Tompkins and Wiehl 1954). But no significant differences were found in skeletal
maturity between Afro- and Euro-American children 4.5 to 14.5 years of age in
Philadelphia (Platt 1956, Bass 1960). The TWI skeletal ages showed little
consistent difference between the Afro- and Euro-American boys, but the Afro-
American girls tended to be somewhat advanced over the Euro-American girls 9 to
12 years of age (Malina 1969, 1970). The NHS data (Roche et al. 1975, 1978)
indicated that the mean skeletal ages tend to be more advanced in the Afro-American
boys than in the Euro-American boys at most ages up to 10 years but not later. The
mean skeletal ages tend to be more advanced in the Afro-American girls than in the
288 CHAPTER 22
The Rehobth Basters (Khoi Khoi x European children) showed the intermediate
TW2 ages between the Kohi Khoi and British children at almost all ages (Singer
and Kimura 1981). According to the preliminary study by the Greulich-Pyle and
TWI methods, the children of mixed Euro-American and Japanese parentage showed
the intermediate maturity between the Japanese in Sapporo and British descendants
in Melbourne (Roche et al. 1971) after pre-adolescence (Kimura 1971, 1972). The
present study confirmed the followings.
1 The children of mixed American and Japanese parentage are followed roughly
a similar pattern of the skeletal maturity with the Japanese.
2 The children of mixed Afro-American and Japanese parentage are always more
advanced in skeletal maturity, followed by those of mixed Euro-American and
Japanese parentage, than the Japanese do.
3 Differences of skeletal maturity between the children of mixed American and
Japanese parentage and the Japanese could be especially due to those for the
RUS bones during preadolescence in both sexes.
4 The children of mixed Afro-American and Japanese parentage are advanced in
skeletal maturity as compared to those of mixed Euro-American and Japanese
parentage, especially for the carpal in the boys and each skeletal age in the
girls.
The Afro-American is advanced in skeletal maturity during adolescence as
compared to the Euro-American, especially in girls (Malina 1969, 1970, Roche et
al. 1975, 1978, Robson et al. 1975, Gam 1976). As the Afro-American is a
European-African admixture, the flow of genes into them has been primarily
undirected with respect to the sex chromosome (European male to African female),
and a greater percentage of the Y chromosome in the Afro-American gene pool are
from European sources (Malina 1970). In this series of mixed American and Japanese
parentage also, a greater percentage of the Y chromosome come from Afro- and Euro-
American males, that is, from European sources.
Examining ossification in children with Klinefelter's and Turner's syndromes,
Tanner et al. (1959) suggested that Y chromosome slows the rate of skeletal
maturation in males. A large majority of studies generally indicated higher
SKELETAL MATURITY IN CHILDREN OF MIXED PARENTAGE 289
5. REFERENCES
Abbott, E.D., Townsen, R.O., French, RB., and Armann, C.F., 1950, Carpal and epiphyseal
development: Another index of nutritional status of rural school children. American Journal of
Disease of Children, 79, 69-81.
Acheson, RM., 1954, A method of assessing skeletal maturity from radiographs. A report from the
Oxford health survey. Journal of Anatomy, 88,498-508.
Acheson, RM., 1960, Effects of nutrition and disease on human growth. In: Tanner, 1.M.(Ed.) Human
Growth. Symposium of the Society for the Study of Human Biology. Vol. 3, pp. 73-92, Pergamon
Press, Oxford.
290 CHAPTER 22
Andersen, E., 1968, Skeletal maturation of Danish school children in relation to height, sexual
development, and social conditions. Acta Paediatrica Scandinavia, Supplement 185.
Andersen, E., 1971, Comparison of Tanner-Whitehouse and Greulich-Pyle methods in a large scale
Danish survey. American Journal of Physical Anthropology, 35, 373-376.
Anzai, I., Sakamoto, K., Togo, M., and Katsunuma, H., 1981, A comparative study of Body composition
of urban and rural Japanese boys 12 to 14 year old. Annals of Human Biology, 8, 109-117.
Ashizawa, K., 1970, Maturation osseuse des enfants japonais de 6 a 18 ans, estimee par la methode de
Tanner-Whitehouse. Bulletins et Memoires de la Societe d'Anthropologie de Paris, 6, 265-280.
Ashizawa, K., 1974, Les erreurs d'interpretation dans I'etude de I'estimation des stades de la
maturation osseuse par un observateur utilisant la methode de Tanner-Whitehouse. Bulletins et
Memoires de la Societe d'Anthropologie de Paris, 13,335-350.
Ashizawa, K., 1993, Skeletal and sexual maturation and growth in Tokyo girls: longitudinal
observations. Clinical Pediatric Endocrinology, 2 (Supplement 1), 5-8.
Ashizawa, K., 1994, Secular trends and regional variations of Japanese skeletal maturation.
Anthropological Science, 102, 193.
Ashizawa, K., Asami, T., Anzo, M., Matsuo, N., Matsuoka, H., Murata, M., Ohtsuki, F., Satoh, M.,
Tanaka, T., Tatara, H., and tsukagoshi, K., 1996, Standard RUS skeletal maturation of Tokyo
children. Annals of Human Biology, 23, 457-469.
Backstrom-Jarvien, L, 1964, Height and weight of Finnish children and young adults. Annals of
Paediartics,23 (Supplement), 1-116.
Bass, W.M., 1960, A comparative study of the sequence and time of appearance of maturity indicators
in the hand in white, Negro and Chinese children in the U.S. American Journal of Physical
Anthropology, n.s.18, 339-340 (abstract).
Baughan, B., Demirjian, A., and Levesque, G-Y., 1979, Skeletal maturity standards for French-
Canadian children of school-age with a discussion of the reliability and validity of such measures.
Human Biology, 51, 353-370.
Beresowski, A., and Lundie, J.K., 1952, Sequence in the time of ossification of the carpal bones in 705
African children from birth to six years of age. South African Journal of Medical Science, 17, 25-
31.
Beunen, G., and Cameron, N., 1980, The reproducibility of TW2 skeletal age assessment by a self-
taught assessor. Annals of Human Biology, 7,155-162.
Beunen, G., Lefevre, J., Ostyn, M., Renson, R., Simons, J., and Van Gerven, D., 1990, Skeletal maturity
in Belgian youths assessed by the Tanner-Whitehouse method (TW2). Annals of Human Biology,
17,355-376.
Bogin, B, 1991, Measurement of growth variability and environment quality in Guatemalan children.
Annals of Human Biology, 18,285-294.
Bogin, B., Sullivan, T., Hauspie, R., and MacVean, R.B., 1989, Longitudinal growth in height, weight,
and bone age of Guatemalan Ladino and Indian schoolchildren. American Journal of Human
Biology, 1, 103-113.
Borms, J., Broekhoff, J., and Beunen, G., 1973, Vergelijkende studie tussen drie methoden voor het
bepalen van de skeletale leeftijd. Archives belges de Medecine sociale, hygiene, Medecine du
Travail et Medecine legale, 3, 343- 353. (cited from Wenzel et al., 1984).
Borns, P.F., and Moshang, T., 1977, Weekly radiology science update. Biometrica, 37, 1-5. Budlinger,
H., and Prader, A, 1972, Unpublished data from Zurich Longitudinal Growth Study, May 1972.
(cited from Eveleth and Tanner, 1976).
Buschke, F., 1935, The radiological examination of skeletons of triplets. Journal of Heredity, 26, 391-
410.
Chang, K.S.F., Chan, S.T., and Low, W.D., 1967, Skeletal maturation of Chinese pre- school children.
Far East Medical Journal, 3, 289-293.
Christie, A., 1949, Prevalence and distribution of ossification centers in the newborn infant. American
Journal of Disease of Children, 77, 355-361.
Dommisse, F.H., and Leipoldt, C.L., 1936, The ossification of the carpal centres in colored children.
South African Medical Journal, 10,713-715.
Douglas, J.W.B., and Blomfield, J.M., 1958, Children Under Five. London, George Allen and Unwin.
Dreizen, S., Snodgrasse, R.M., Parker, a.s., Currie, c., and Spies, T.D., 1954, Maturation of bone
centers in hand and wrist of children with chronic nutritive failure. American Journal of Disease of
Children, 87,429-439.
SKELETAL MATURITY IN CHILDREN OF MIXED PARENTAGE 291
Dreizen, S., Snodgrasse, R.M., Webb-Peploe, H., and Spies, T.D., 1957, The effect of prolonged
nutritive failure on epiphyseal fusion in the human hand skeleton. American Journal of
Roentgenolgy, Radium Therapy and Nuclear Medicine, 78,461-470.
Dreizen, S., Snodgrasse, R.M., Webb-Peploe, H., and Spies, T.D., 1958, The retarding effect of
protracted undernutrition on the appearance of the postnatal ossification centers in the hand and
wrist. Human Biology, 30, 253-264.
Dreizen, S., Spirakis, C.N., and Stone, R.E., 1964, Chronic undernutrition and postnatal ossification.
American Journal of Disease of Children, 108,44-52.
Dreizen, S., Spirakis, C.N., and Stone, R.E., 1967, A comparison of skeletal growth and maturation in
undernourished and well-nourished girls before and after menarche. Journal of Pediatrics, 70, 156-
263.
Eto, M., 1971, The skeletal development of the children in Tokyo -Comparative study Through two
kinds of the skeletal age assessed by Japanese and British Methods -. Journal of Anthropological
Society of Nippon, 79, 9-20.
Eto, M., and Ashizawa, K., 1992, Longitudinal Observations on Physical Growth and TW2 Skeletal
Maturation of Girls in Tokyo. Radiographic Atlas of Hand and Wrist. Tokyo, Terpeia.
Eveleth, P.B., 1978a, Population differences in growth: environmental and genetic factors. In: Falkner,
F. and Tanner, lM. (Eds.), Human Growth, Vol. 3 (Rev. Ed.), New York, Plenum Press, pp. 373-
394.
Eveleth, P.B., 1978b, Differences between populations in body shape of children and adolescents.
American Journal of Physical Anthropology, 49, 373-382.
Eveleth, P.B., and Tanner, lM., 1976, Worldwide Variations in Human Growth. Cambridge, Cambridge
University Press, pp. 199-206.
Falkner, F., 1958, Skeletal maturation; an appraisal of concept and method. American Journal of
Physical Anthropology, 16,381-396.
Fry, E.I., 1966, Skeletal age by guess and by Tanner II. Large sample analysis. American Journal of
Physical Anthropology, 25, 202. (quoted from Roche et aI., 1974).
Garn, S.M., 1976, Metacarpal Lengths, Cortical Diameters and Areas from the 10-State Nutrition
Survey, Including Estimated Skeletal Weights, Height, and Stature For Whites, Blacks and
Mexican-Americans. Publication No. (HSM) 21-72-522. U.S. Department of Health, Education and
Welfare, Washington, D.C.
Garn, S.M., and Clark, D.C., 1975, Nutrition, growth, development, and maturation findings from the
ten-state nutrition survey of 1968-1970. Pediatrics, 56, 306- 319.
Garn, S.M., and McCreery, L.D., 1970, Variability of postnatal ossification timing and evidence for a
"dosage" effect. American Journal of Physical Anthropology, 32, 139-144.
Garn, S.M., and Rohmann, C.G., 1962a, Parent-child similarities in hand-wrist ossification. American
Journal of Disease of Children, 103,603-607.
Gam, S.M., and Rohmann, C.G., 1962b, X-linked inheritance of developmental timing in man. Nature,
196, 695-696.
Garn, S.M., Rohmann, C.G., and Apfelbaum, B., 1961a, Complete epiphyseal union of the hand.
American Journal of Physical Anthropology, 19,365-372.
Garn, S.M., Rohmann, C.G., and Davis, A., 1963, Genetics of hand-wrist ossification. American Journal
of Physical Anthropology, 21, 33-40.
Garn, S.M., Rohmann, C.G., and Hertzog, K.P., 1969, Apparent influence of the X chromosome on
timing of73 ossification centers. American Journal of Physical Anthropology, 30, 123-128.
Gam, S.M., Rohmann, C.G., and Robinow, M., 1961b, Increments in hand-wrist ossification. American
Journal of Physical Anthropology, 19,45-53.
Goldstein, H., 1971, Factors influencing the height of seven-year-old children. Results from the
National Child Development Study. Human Biology, 43, 92-111.
Goto, A., 1961, Studies on the color tone and the histological structure of the skin of mixed blood.
Journal of Nippon Medical School, 28, 145-158. (in Japanese).
Gray, S., and Lamons, F., 1959, Skeletal development and tooth eruption in Atlanta children. American
Journal ofOrthod., 45, 272-277.
Greulich, W.W., and Pyle, S.L., 1950, Radiographic Atlas of Skeletal Development of the Hand and
Wrist. Stanford, Stanford University Press.
Greulich, W.W., and Pyle, S.L., 1959, Radiographic Atlas of Skeletal Development of the Hand and
Wrist. 2nd Ed. Stanford, Stanford University Press.
292 CHAPTER 22
Guilfored, M.C., Chinn, S., and Rona, R.I., 1991, Social environment and height: England and Scotland
1987 and 1988. Archives of Diseases in Childhood, 66, 235-240.
Habicht, J-P., Yarbrough, C., Martorell, R., Malina, R.M., and Klein, R.E., 1974, Height and weight
standards for preschool children. Lancet, I, 611-615.
Hamill, P.V.V., Johnston, F.E., and Lemeshow, S, 1972, Height and Weight of Children; Socioeconomic
Status. Department of Health, Education and Welfare, Pub. No. (HSM) 73-1601. Vital Health
Statitics Series 11, No. 11. Washington, D.C.
Hanihara, K., 1954-57, Studies on the deciduous dentition of the Japanese and the Japanese-American
hybrids. I. Journal of Anthropological Society of Nippon, 63, 165-185; 64, 63-82, 95-116; 65, 67-78,
151-164.
Hanihara, K., 1963, Crown characters of the deciduous dentition of the Japanese and the Japanese-
American hybrids. Dental Anthropology, pp. 105-124, Pergamon, Oxford.
Hanihara, K., 1965, Some crown characteristics of the deciduous incisors and canines in Japanese-
American hybrids. Journal of Anthropological Society of Nippon, 72, 135-145.
Hanihara, K., 1968, Morphological patterns of the deciduous dentition in the Japanese- American
hybrids. Journal of Anthropological Society of Nippon, 76, 114-121.
Helm, S., 1979, Skeletal maturity in Danish schoolchildren assessed by TW2 method. American Journal
of Physical Anthropology, 51, 345-352.
Hertzog, K.P., Falkner, F., and Garn, S.M., 1969, The genetic determination of Ossification sequence
polymorphism. American Journal of Physical Anthropology, 30, 141-144.
Hess, A.F., and Abramson, H, 1933, Familial retardation in ossification of the carpal centers. Journal of
Pediatrics, 3, 158-165.
Hess, A.F., and Weinstock, M., 1925, A comparison of the evolution of carpal centres in white and
Negro newborn infants. American Journal of Disease of Children, 29, 347-354.
Himes, J.H., 1984, An early hand-wrist atlas and its implications for secular change in bone age. Annals
of Human Biology, 11,71-75.
Hojo, T., 1964, Untersuchungen tiber die Mundzone bei den Japaner-Americaner- Mischlingen.
Journal of Anthropological Society of Nippon, 71, 143-152. (in Japanese with German summary).
Honda, A., 1958, On the iris and head-hair of the half-breed children. Journal of Nippon Medical
School, 25, 297-322.
Hoshi, H., 1959, Brief note on some cephalometrical results from the longitudinal study of the Japanese-
American hybrids. Journal of Anthropological Society of Nippon, 67, 73-81.
Hoshi, H., 1970, Physical growth of Japanese-American hybrids from 6 to 15 years of age with special
reference to genetic-environmental relationships. Monograph of Society of Research of Child
Development, 35, 49-54.
lshimoto, G., 1968, Urinary excretion of Ei,-aminoisobutyric acid in Japanese- American hybrids.
Journal of Anthropological Society of Nippon, 68, 141-144.
Johnston, F.E., 1962, Longitudinal Study of Skeletal Maturation and Its Relationship to Growth in
Philadelphia White Children; a Dissertation in Anthropology, Unpublished doctoral dissertation,
University of Pennsylvania. (quoted from Roche et at., 1974).
Johnston, F.E., 1963, Skeletal age and its prediction in Philadelphia children. Human Biology, 35, 192-
202.
Johnston, F.E., and Jahina, S.B., 1965, The contribution of the carpal bones to the assessment of skeletal
age. American Journal of Physical Anthropology, 23, 349-354.
Jones, D.L., Hempill, W., and Meyer, E.SA, 1973, Height, Weight and Other Physical Characteristics
of New South Wales Children. Part I. Children Aged Five Years and Over. Sydney, N.S.W.
Department of Health. (quoted from Eveleth, 1978).
Jones, P.R.M., and Dean, R.F.A., 1956, The effects of kwashiorkor on the development of the bones of
the hand. Journal of Tropical Medicine, 2, 51-68.
Kariberg, P., Taranger, J., EngstrOm, I., Lichtenstein, H., and Svenberg-Redegren, I., 1976, The
somatic development of children in a Swedish urban community. A prospective longitudinal study.
Acta Paediatrica Scandinavica (Supplement 258).
Kawashima, M., Nanjo, F., Sakura, H., and Ohta, S., 1972, Skeletal age estimated by the Tanner-
Whitehouse-Heary method. Comparison with the Sugiura and Nakazawa method. Clinical
Orthopedic Surgery, 7, 350-358. (in Japanese).
Kelly, H.I. and Reynolds, L., 1947, Appearance and growth of ossification centers and increases in the
body dimensions of white and Negro infants. American Journal of Roentgenology and Radium
Therapy, 57,477-516.
SKELETAL MATURITY IN CHILDREN OF MIXED PARENTAGE 293
Key, W.E., 1936, Radiological studies of the skeleton of twins, triplets and quadruplets. Journal of
Heredity, 27,86-88.
Kimura, K., 1962, Radiological studies of the hand and wrist bones in twins. In: Fujita, T. (Ed.), The
Study of Twins-III. Tokyo, Maruzen, pp. 44-56. (in Japanese).
Kimura, K., 1971, On the skeletal maturation of the Japanese-American hybrids- A preliminary report.
Journal of Anthropological Society of Nippon, 79, 21-29.
Kimura, K., 1972, Skeletal maturation in Japanese according to the Oxford and Tanner- Whitehouse
methods. Acta Anatomica Nipponica, 47, 358-372.
Kimura, K., 1974a, An application of the method of biological distance to the dermatoglyphic data of
Ainu-Japanese and Japanese American white hybrids. Journal of Anthropological Society of
Nippon, 13,37-50.
Kimura, K., 1974b, Dermatoglyphics of the Japanese-American hybrids. Bulletin of Faculty of Physical
Education, Tokyo University of Education, 13,37-50.
Kimura, K., 1976a, Skeletal maturation of Japanese-American white hybrids. American Journal of
Physical Anthropology, 44, 353-356.
Kimura, K., 1976b, Skeletal maturation of children in Okinawa. Annals of Human Biology, 3, 149-155.
Kimura, K., 1977, Skeletal maturity of the hand and wrist in Japanese children in Sapporo by the TW2
method. Annals of Human Biology, 4, 353-356.
Kimura, K., 1981, Skeletal maturity in twins. Journal of Anthropological Society of Nippon, 89, 457-
478.
Kimura, K., 1983, Skeletal maturity and bone growth in twins. American Journal of Physical
Anthropology, 60, 491-497.
Kimura, K., 1984, Has the secular trend stopped in Japanese? Japanese Journal of Human Posture, 4,
17-27. (in Japanese with English summary).
Kimura, K., 1994, Skeletal maturation of American-Japanese hybrids by TW2 method. Anthropological
Science, 102, 194.
Kimura, K., 1996, Growth of second metacarpal of American-Japanese hybrids. Acta Anatomica
Nipponica, 71, 39-46.
Kimura, K., 1999, Skeletal maturity and growth of children in two sections of Tokyo in the 1960s,
Perspective in Human Biology, 4,181-187.
Kimura, K., and Takeuchi, K., 1982, Notches at the n for the carpal age on-epiphyseal end of the
second metacarpal in Japanese-American hybrids. Zeitschrift fUr Morphologie und Anthropoiogie,
73, 1-14.
Kitamura, Sh., and Suzuki, K., 1965, Blood groups of the hybrid children. III. subgroups of A and E-e
blood factor. Tokyo Journal of Medical Science, 64, 171-178.
Kohara, Y., 1968, Subcutaneous fat distribution in Japanese-American hybrid boys. Journal of
Anthropological Society of Nippon, 76, 105-113.
Kondo, S, and Eto, M., 1972, Physical growth studies on Japanese-American hybrids in comparison with
native Japanese. In: Proceedings of Meeting for Review and Seminar of the US-Japan Cooperative
Research on Human Adaptability, Kyoto, 1972, Japan Society for the Promotion of Science and
National Sciences Foundation, Kyoto.
Kondo, S., Sato, M., Kikuchi, Y., Tomita, M., Okada, M., and Hayami, A., 1968, Electoromyographic
studies on facial muscles of Japanese-American hybrids. Journal of Anthropological Society of
Nippon, 76, 135-140.
Koski, K., Haataja, 1., and Lappalanien, M., 1961, Skeletal development of hand and wrist in Finnish
children. American Journal of Physical Anthropology, 19,379- 382.
Kristmundsdottir, F., Burwell, R.G., Marshall, W.A., and Small, P., 1984, Cross- sectional study of
skeletal maturation in normal children from Nottingham and London. Annals of Human Biology, II,
133-139.
Laska-Mierzejewska, T., Milicer, H., and Piechaczek, H, 1982, Age at menarche and its secular trend
in urban and rural girls in Poland. Annals of Human Biology, 9, 284-292.
Levine, E., 1972, The skeletal development of children of four South African populations. Human
Biology, 44, 399-412.
Lindgren, G., 1976, Height, weight and menarche in Swedish urban school children in relation of
socioeconomic and regional factors. Annals of Human Biology, 3,501-528.
Lindgren, G., and Cernerud, L., 1992, Physical growth and socioeconomic background of Stockholm
schoolchildren born in 1933-63. Annals of Human Biology, 19, 1-16.
294 CHAPTER 22
Low, W.O., 1972, Relation between socio-economic status and skeletal maturation of Chinese Children.
Zeitschrift fur Morphologie und Anthropologie, 64, 1-11.
Low, W.O., Chan, S.T., Chang, K.S.F., and Lee, M.M.C., 1964, Skeletal maturation of Southern
Chinese. Child Development, 35, 1313-1336.
Low, W.O., Fung, S.H., and Cerny, E.B., 1976a, Biological age of Southern Chinese boys. Zeitschrift
fur Morphologie und Anthropologie, 67, 136-144.
Low, W.O., Fung, S.H., and Cerny, E.E., 1976b, Biological age of Southern Chinese girls. Zeitschriffilr
Morphologie und Anthropologie, 67, 264-274.
Mackay, D.H., 1952, Skeletal maturation in the hand: A study of development in East African children.
Trans. Royal Society of Tropical Medicine and Hygiene, 46,135-150.
Mackenbach, J.P., 1991, Narrowing inequalities in children's height, Lancet, 338, 764.
Malina, R.M., 1969, Skeletal maturation rate in North American Negro and White children. Nature,
223,1075.
Malina, R.M., 1970, Skeletal maturation studied longitudinally over one year in American whites and
Negroes six through thirteen years of age. Human Biology, 42, 377-390.
Maniar, B.M., Seevai, M.H., and Kapur, P.L., 1974, A study of ossification centers in the hand and wrist
ofIndian children. Indian Pediatrics, 11, 203-211.
Marshall, W.A., Ashcroft, M.T., and Bryan, G., 1970, Skeletal maturation of the hand and wrist of
Jamaican children. Human Biology, 42, 419-435.
Masse, G., and Hunt, E.B. Jr., 1963, Skeletal maturation of the hand and wrist in West African children.
Human Biology, 42, 419-435.
Mathiasen, M.S., 1973, Determination of bone age and recording of minor skeletal hand anomalies in
normal children. Danish Medical Bulletin, 20, 80-85. (quoted from Eveleth and Tanner, 1976).
Matsuzawa, Sh., and Kitamura, Sh., 1956, Blood groups of the hybrid children. I. ABO Groups, MN
types, Rh types, secretors and nonsecretors. Tokyo Journal of Medical Science, 64, 83-88.
Medicus, H., Gron, A., and Moorrees, C.F.A., 1971, Reproducibility of rating stages of osseous
development (Tanner-Whitehouse system). American Journal of Physical Anthropology, 35, 359-
371.
Meredith, H.V., 1979, Comparative findings on body-size of children and youth living at urban centers
and rural areas. Growth, 43, 95-104.
Meredith, H.V., 1984, Body size of infants and children around the world in relation to socioeconomic
status. Archives of Child Development and Behavior, 18,81- 145.
Michaut, E., Niang, J., and Dan, V., 1972, La maturation osseous pendant la peri ode pubertaire.
Annales de Radiologie, 15,767-779. (quoted from Eveleth and Tanner, 1976).
Michibe, T., 1961, On the skin color of the half-breed children. Journal of Nippon Medical School, 28,
1196-1223. (in Japanese).
Omoto, K., 1968, Studies of skin color among hybrids subjects in Japan with special reference to
measurements of Sun-tanning. Journal of Anthropological Society of Nippon, 76,122-134.
Otto, U, 1976, Male youth. A socio-psychiatric study of a total annual population of Swedish adolescent
boys. Acta Psyciatrica Scandinavia, 264 (Supplement), 1- 312.
Platt, R.A., 1956, The Skeletal Maturation of Negro School Children. Unpublished master's thesis.
University of Pennsylvania, Philadelphia. (quoted from Malina, 1970).
Poland, J., 1898, Skiagraphic Atlas Showing the Development of the Bones of the Wrist and Hand.
London, Smith, Elder & Co., pp. 41. (quoted from Himes, 1984).
Prader, A., Tanner, lM., and von Harnack, G.A., 1963, Catch-up growth following illness or starvation.
An example of developmental canalization in man. Journal of Pediatrics. 62, 646-659.
Prado, C., Andersen, A.H., and Martinez, R., 1986, The relationship obesity-blood pressure: Attending
to ontogenesis and socio-environmental variation. International Journal of Anthropology (Spain), 1,
153-164.
Prahl-Andersen, B., Kowalski C.J., and Heydendael, P.H.lM., 1979, A Mixed- Longitudinal
Interdisciplinary Study of Growth and Development. London, Academic Press. (quoted from
Wenzel et al., 1984).
Preece, M.A., 1989, Standardization of growth. Acta Paediatrica Scandinavica, 349, 57- 64.
Pryor, J.W., 1907, The hereditary nature of variation in the ossification of bones. Anatomical Record, 1,
84-88.
Pryor, J.W., 1936, Ossification as additional evidence in differentiating identicals and fraternals in
mUltiple births. American Journal of Anatomy, 59, 403-409.
Pryor, lW., 1939, Normal variations of in the ossification of bones. Journal of Heredity, 30, 249-255.
SKELETAL MATURITY IN CHILDREN OF MIXED PARENTAGE 295
Reynolds, E.L., 1943, Degree of kinship and pattern of ossification. American Journal of Physical
Anthropology, 1,405-416.
Rigler, L.G., 1938, Roentgen studies of twins and triplets. Radiology, 30, 461-470.
Robson, J.R.K., Larkin, F.A., Bursick, J.H., and Perri, K.P., 1975, Growth standards for Infants and
children: a cross-sectional study. Pediatrics, 56, 1014-1020.
Roche, A.F., 1967, A study of skeletal maturation in a group of Melbourne children. Australian
Paediatric Journal, 3,123-127.
Roche, A.F., 1989, Relative utility of carpal skeletal age. American Journal of Human Biology, 1,479-
482.
Roche, A.F., and Davila, G.H., 1976, The reliability of assessments of the maturity of individual hand-
wrist bones. Human Biology, 48, 585-597.
Roche, A.F., Davila, G.H., and Eyman, S.L., 1971, A comparison between Greulich- Pyle and Tanner-
Whitehouse assessments of skeletal maturity. Radiology, 98, 273-280.
Roche, A.F., Roberts, J., and Hamill, P.V.V., 1974, Skeletal Maturity of Children 6-11 Years. Vital
Health Statistics-Series II-No. 140, DHEW Publication No. (HRA) 75-1622. U.S. Department of
Health, Education and Welfare, Washington, D.C., pp. 62.
Roche, A.F., Roberts, J., and Hamill, P.V.V., 1976, Skeletal Maturity of Youths 12-17 Years. Vital
Health Statistics-Series l1-No. 160, DHEW Publication No. (HRA) 77-1642. U.S. Department of
Health, Education and Welfare, Washington, D.C., pp. 90.
Roche, A.F., Roberts, J., and Hamill, P.V.V., 1975, Skeletal maturity of youths 6-11 years: racial,
geographic area and socioeconomic differentials. Vital Health Statistics 11 - No. 149, DHEW
Publication No. (PHS) 76-1631. U.S. Department of Health Education and Welfare, Washington,
D.C., pp. 81.
Roche, A.F., Roberts, J., and Hamill, P.V.V., 1978, Skeletal Maturity of Youths 12-17 Years: Racial,
Geographic Area and Socioeconomic Differentials. Vital Health Statistics-Series II-No. 167,
DHEW Publication No. (PHS) 79-1654. U.S. Department of Health, Education and Welfare,
Washington, D.C., pp. 98.
Roede, M., Stoelinga, G., De Jong-van der Kar, M., and van 't Hof, M., 1976, Het Nijmeegse
Groeionderzoek. Antropometrische gegevens betreffende gezonde Nederlandse kinderen van 5
112, 8 1/2 en 10 1/2 jaar. Nederlands Tijdschrift voor Geneeskunde, 120, 133-140. (quoted from
Wenzel et al., 1984).
Rosique, J., and Rebato, E., 1995, Comparative study of statural growth in Spanish population. American
Journal of Human Biology, 7, 553-564.
Sahkir, A., and Zini, S., 1974, Skeletal maturation of the hand and wrist of young children in Baghdad.
Annals of Human Biology, 1, 189-199.
SAS Institute Incorporation, 1958, SAS User's Guide: Statistics. 5th Ed. Cary, North Carolina, SAS
Institute Incorporation.
Satoh, M., Tanaka, 1., Ashizawa, K., Ohtuski, F., Anzo, M., Matsuo, N., Tatara, H., and Murata, M.,
1993, Inter- and intra-observer errors on estimation of the bone age by Tanner-Whitehouse II
method. Clinical Pediatric Endocrinology, 2,1055-1059.
Singer, R., and Kimura, K., 1981, Body height, weigh, and skeletal maturation in Hottentot (Khoikhoi)
children. American Journal of Physical Anthropology, 54, 401-413.
Sklad, V.M., 1973, Genetische Grundlagen einiger mit dem Heranreifen verbundener Erscheinungen
(nach Awillingsuntersuchungen). Zeitschrift fOr Morphologie und Anthropolgie, 65, 192-211.
Snedecor, G.W., and Cochran, W.G., 1980, Statistical Methods (9th Ed.). Ames, The Iowa State
University Press, pp. 243-248. (quoted from Roche and Guo, 1989).
Stinson, S., 1985, Sex differences in environmental sensitivity during growth and development.
Yearbook of Physical Anthropology, 28, 123-147.
Stuart, H.C., Pyle, S.l., Cornoni, I., and Reed, R.B., 1962, Onsets completions and spans of ossification in
the 29 bone-growth centers of the hand and wrist. Pediatrics, 29, 237-249.
Suda, A., 1968, Introduction to our anthropological study on the Japanese-American hybrids. Journal of
Anthropological Society of Nippon, 76, 89-94.
Suda, A., Hoshi, H., Eto, M., Ashizawa, K., 1975, Longitudinal observation on the pectoral growth of
Japanese-American hybrids with special reference to the difference in growth pattern between
thoracal and extrathoracal parts. Journal of Anthropological Society of Nippon, 83, 95-104.
Suda, A., Hoshi, H., Eto, M., Ashizawa, K., 1976a, Longitudinal studies on the Japanese-American
hybrids. Anthropological and Genetical Studies On the Japanese. JIBP Synthesis-Human
Adaptability, Vol. 2. Tokyo, University of Tokyo Press, pp. 33-72.
296 CHAPTER 22
Suda, A, Hoshi, H., Eto, M., Ashizawa, K., 1976b, Longitudinal observation on the limb growth of
Japanese-American hybrids with special regard to seasonal variation. Journal of Anthropological
Society of Nippon, 84, 15-30.
Suda, A., Hoshi, H., Eto, M., Ashizawa, K., and Hojo, Teruyuki, 1973, Longitudinal observation on
anterior trunk height, iliospinal height, biacromial breadth and bicristal breadth of Japanese-
American hybrids from 6 to 15 years of age. Journal of Anthropological Society of Nippon, 81,
185-194.
Suda, A, Hoshi, H., Sato, M., Eto, M., and Ashizawa, K., 1968, Longitudinal observation on the chest
circumference and sitting height of Japanese-American hybrids from 6 to 15 years of age. Journal
of Anthropological Society of Nippon, 76, 95-104.
Suda, A, Yamaguchi, 8., Hoshi, H., Endo, B., and Eto, M., 1965, Longitudinal observation on the
stature and body weight of Japanese-American hybrids from 6 to 15 years of age. Journal of
Anthropological Society of Nippon, 73, 54-63.
Suzuki, K., and Matsuzawa, Sh., 1956, Blood groups of the hybrid children. II. P (Q) groups. Tokyo
Journal of Medical Science, 64, 105-108.
Takai, S., 1990, Smoothed skeletal maturity curve of Japanese children by Tanner- Whitehouse (TW2)
method and its application. Acta Anatomica Nipponica, 65, 436-447.(in Japanese with English
summary).
Takai, S., 1993, Velocities for the Tanner-Whitehouse 2 skeletal maturity in northwestern Japanese
children. Okajimas Folia Anatomica Japonica, 70, 119- 126.
Takai, S., 1996, Does early skeletal maturer tend to be a smaller adult? Bulletin of Institute of Health
and Sport Sciences, University of Tsukuba, 19,91-97.
Takai, S., Akiyoshi, T., and Fuchigami, A, 1984, Skeletal maturity of Japanese children in Amami-
Oshima Island. Annals of Human Biology, 11,571-575.
Takai, S., and Akiyoshi, T., 1983, Skeletal maturity of Japanese children in western Kyushu. American
Journal of Physical Anthropology, 62, 199-204.
Tanner, J.M., and Gibson, R.D., 1994, A computerized image analysis system for estimating Tanner-
Whitehouse 2 bone age. Hormone Research, 42, 282-287.
Tanner, J.M., Oshman, D., Lindgren, G., Grunbaum, J.A., Elsouki, R., and Labarthe, D., 1994,
Reliability and validity of computer-assessed estimates of Tanner- Whitehouse skeletal maturity
(CASAS): comparison with manual method. Hormone Research, 42, 288-294.
Tanner, J.M., Prader, A., Habicht, H., and Ferguson-Smith, M.A, 1959, Genes on the Y chromosome
influencing rate of maturation in man. Lancet, 2, 141-144.
Tanner, J.M., Whitehouse, R.H., and Healy, M.J.R., 1962, A New System for Estimating Skeletal
Maturity from the Hand and Wrist, with Standards derived from a Study of 2,600 Healthy British
Children. Paris, International Children's Center.
Tanner, J.M., Whitehouse, R.H., Cameron, N., Marshall, W.A, Healy, MJ.R., and Goldstein, H., 1983,
Assessment of Skeletal Maturity and Prediction of Adult Height (TW2 method). 2nd Ed. London,
Academic Press.
Taranger, J., Karlberg, J., Bruning, B., and EngstrOm, I., 1987, Standard deviation score chart of
skeletal maturity and its velocity in Swedish children assessed by the Tanner-Whitehouse method
(TW2). Annals of Human Biology, 14,357-365.
Tatara, H., Murata, M., Takaishi, M., and Funakawa, H., 1985, Estimation of bone age of Japanese
children, especially based on longitudinal data. Pediatrics, 26, 455- 459. (in Japanese).
Tiisala, R., Kantero, R., and Tamminen, T., 1971, A mixed longitudinal study of skeletal maturation in
healthy Finnish children aged 5 to 10 years. Human Biology, 43, 224-236.
Tiisala, R., Kantero, R., BackstrOm and Hallman, K., and Hallman, N., 1969, A mixed longitudinal study
on skeletal maturation in healthy Finnish children aged 5 to 10 years. Human Biology, 41, 560-579.
Todd, T.W., 1931, Differential skeletal maturation in relation to sex, race, variability and disease. Child
Development, 2, 49-65.
Todd, T.W., 1937, Atlas of Skeletal Maturation (Hand). St. Louis, C.V. Mosby Co. Tompkins, W.T., and
Wiehl, D.G., 1954, Epiphyseal maturation in the newborn as related to maternal nutritional status.
American Journal of Obstetrics and Gynecology, 68, 1366-1377.
Valaoras, V., and Laros, K., 1969, Biometric characteristics of Greek pupils in elementary Schools.
IATPKH, 15,266-276. (quoted from Eveleth, 1979).
Van Loon, H., Saverys, V., Vuylsteke, J.P., Vlietinck, R.F., and Eeckels, R., 1986, Local versus
universal growth standards: The effect of using NCHS as a universal reference. Annals of Human
Biology, 13,347-357.
SKELETAL MATURITY IN CHILDREN OF MIXED PARENTAGE 297
Venrooij-Ysselmuiden, M.E., van, and Ipenburg, A.V., 1978, Mixed longitudinal data on skeletal age
from a group of Dutch children living in Utrecht and surroundings. Annals of Human Biology, 5,
359-380.
Vignolo, M., Milani, S., Cerbello, G., Coroli, P., Di Batista, E., and Aicardi, G., 1992, Fels, Greulich-
Pyle, and Tanner-Whitehouse bone age assessments in a group of Italian children and adolescents.
American Journal of Human Biology, 4, 493-500.
Waldmann, E., Baber, F.M., Field, C.E., Billewicz, W.l., and Thomson, A.M., 1977, Skeletal maturation
of Hong Kong Chinese children in the first five years oflife. Annals of Human Biology, 4, 343-352.
Weiner, J.S., and Thambipillai, V., 1952, Skeletal maturation of West African Negroes. American
Journal of Physical Anthropology, 10,407-418.
Wenzel, A., and Melsen, B., 1982, Skeletal maturity in 6-16-year-old Danish children assessed by the
Tanner-Whitehouse method. Annals of Human Biology, 9, 277- 281.
Wenzel, A., Dorschl, H., and Melsen, B., 1984, Skeletal maturity in Austrian children assessed by the
GP and TW-2 methods. Annals of Human Biology, II, 173- 177.
Wetherington, R., and Rutenber, G.W., 1978, Contribution to variance in phalangeal growth: estimates
from twins. American Journal of Physical Anthropology, 48,83-88.
Wingerd, J., Peritz, E., and Sproul, A., 1974, Race and stature differences in the skeletal maturation of
the hand and wrist. Annals of Human Biology, 1: 201-209.
Xi, H., Roche, A.F., and Guo, S., 1989, Sibling correlations for skeletal age assessments by the Fels
method. American Journal of Human Biology, 1,613-619.
Ye, Y., Wang, C., and Cao, L., 1992, Skeletal maturity of the hand and wrist in Chinese children in
Changsha assessed by TW2 method. Annals of Human Biology, 19,427-430.
Yoshida, N., 1997, Bone age assessed by computer-assessed skeletal age (CASA) system: comparison
with the manual method. Journal of Osaka Medical College, 55, 43-54. (in Japanese with English
summary).
Zhen, 0., and Baolin, L., 1986, Skeletal maturity of the hand and wrist in Chinese school children in
Harbin assessed by the TW2 method. Annals of Human Biology, 13, 183-187.
BIOLOGICAL ASPECTS OF GROWTH
CHAPTER 23
O. G. EIBEN, A. NEMETH
Department of Biological Anthropology, E6tv6s Lorimd University, Budapest,
Hungary
1. INTRODUCTION
Growth studies in Hungary have been carried out since the 1870s (Eiben 1988a).
The authors of that time were usually content with measuring stature. Systematic
growth studies with remarkable anthropometric programmes started after World War
II, and included amongst others, in the 1980s the Hungarian National Growth Study
(thereafter HNGS, Eiben et al. 1991). Before and after this, during the 20th Century,
there were many regional studies, too.
Growth of children and youth has been recommended as one of the best indices of
health and nutritional status of a community (WHO 1976). Starting out from the
scientifically established fact that growth and development of children and youth
monitors the biological status of the whole population (Tanner 1978, Bielicki 1986,
Eiben 1988b), and having a sense of responsibility for Hungarian children's welfare,
Eiben, with his colleagues A. Barabas and E. Pant6, organised and carried out the
above mentioned study.
The HNGS was a nation-wide, representative cross-sectional growth and physical
fitness study. The aim of it was to analyse growth and maturation as well as
physical fitness of Hungarian children and youth, taking into consideration all the
ecological factors existing in Hungary in the 1980s, especially regrouping of the
population, urbanisation, the urban and rural mode of life, etc. In the last third of our
20th Century, namely, these social factors were the most important environmental
ones influencing growth and maturation process.
The first HNGS was planned in 1981 and was designed to meet two important
needs:
• To provide normative standards to assess and monitor individual growth and
physical fitness, and
• to establish a baseline for successive sampling studies to analyse changes in the
Hungarian population.
Field studies commenced in January 1982 and were completed by March 1985 (see
Eiben et al. 1991).
The investigators of the HNGS visited 113 settlements in Hungary, more than
350 nurseries and schools. Their sample contained 39,035 healthy boys and girls
301
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 301-312.
© 2001 Kluwer Academic Publishers.
302 CHAPTER 23
between 3 and 18 years of age. This sample represents 1.5% of the same aged
Hungarian youth.
The anthropometric programme contained 18 body measurements. Seven
motoric tests were taken and socio-demographic data of the family background were
collected. Tools and techniques of the investigations were according to the
internationally standards (Martin and Saller 1957, Tanner et al. 1969, AAHPER
1965, Matthews 1973, Haag and Dassel 1975, Simon and Renson 1982)
The fIrst Hungarian national growth standards for 18 body measurements and 7
motoric tests were published in the middle 1980s (Eiben and Pant6 1986, 1987/88,
Eiben et al. 1991). This reference data serve as an 'etalon' in Hungary for growth
and physical fItness studies. Ten years after initial the HNGS, we were able to repeat
our study in Budapest, in the capital of Hungary.
3. RESULTS
Tables are shown in the Appendix. Budapest children in the middle of the 1990s
show a slight positive secular trend compared to the HNGS (Nemeth and Eiben
1997). Mean height and body mass of boys' and girls' age groups investigated are
SOMATOTYPES OF BUDAPEST CHILDREN 303
~ud.pest boys
I
MESOVORPHY
~
Z
D
o
~
cr:
3: o
o::0 :::<:
o
/~ t> .........
UJ
Figure 1. Changes in somatotype of Budapest boys between the age of 3 and 18 years.
~cr:
Z
o
g,
o
3:
o o
:;0 :::<:
/~ vL.LI
Figure 2. Changes in somatotype of Budapest girls between the age of 3 and 18 years.
SOMATOTYPES OF BUDAPEST CHILDREN 305
4. DISCUSSION
Carter and Heath based their statement on several somatotype studies, carried out
all over the world, in which the authors investigated children and adolescents and in
some cases also followed their changes in somatotype with age. It is good form to
mention first the Atlas of Tanner and Whitehouse (1982), based on their famous
Harpenden Growth Study. Further papers using Heath - Carter method are at our
disposal (in Belgium: Duquet 1980; in Bohemia and Moravia: Stepnicka 1976; in
Hungary: Farmosi 1982, Eiben (Kormend) 1985, Bodzsar (Bakony) 1982, 1991; in
Finland: Holopainen et al. 1984; in India (Gaddi, Raj put): Singh and Sidhu 1980,
(Jat Sikh Punjab) Kansal 1981, (Bania, Punjab) Kansal 1981, (Bangalore) Rangan
1982; in Venezuela (Caracas) Perez et al. 1985; in Nigeria (lle-Ife) Toriola and
Igbokwe 1985; in Brazil (Londrina) Guedes 1983; in the U.S.A. (Illinois) Slaughter
et al. 1980, etc. - see all these: Carter and Heath 1990 pages 160-164).
It is really illuminating to follow the wandering pattern of the mean changes in
somatotypes with age. We can demonstrate this with some Hungarian examples.
Eiben (1977) found in his Western Hungarian "Kormend Growth Study" of 1968, in
contrast to the present sample, that mean somatotypes in early childhood in both
sexes could be found in the ecto-mesomorphic field. In boys, mean somatotypes
shifted with age through the central field to the meso-ectomorphy, while in girls,
these means moved through the central field to the meso-endomorphic region. Ten
years later, in his 1978 Kormend Growth Study, Eiben found a striking difference:
endomorphy was far higher in both genders than in his previous study (Eiben 1985,
1988b). These later findings are more similar to those of our recent Budapest
sample.
Szmodis (1977) examined Hungarian boys and girls aged 5-17 years, among
them young athletes as well. Those boys were more ectomorphic and girls were less
endomorphic. Even mean somatotypes of the older girls in his sample were in the
central field of the somatochart.
Bodzsar (1982) investigated 10-14 year old Hungarian girls, and she found
central mean somatotypes, and increase of endomorphy and decrease of mesomorphy
with age. The same author published data of 6.5-14.5 year-old Hungarian boys and
girls in her other study (Bodzsar 1991). Mean somatotypes in this study differ
significantly from those of our recent sample. Somatotypes of boys in Bodzsar's
study moved from the ecto-mesomorphic field to the meso-ectomorphic one with
increasing age. Values of girls shifted from the ecto-mesomorphic filed through the
central field to the ecto-endomorphic region. According to this, endomorphy in
general was lower and ectomorphy in general was higher in Bodzsar's sample. The
only exception was that in 13-14 year-old girls in endomorphy were more or less
equal with that of our current sample.
Our Budapest study, presented here, provides convincing evidence that
somatotype analysis results in adequate information about changes in physique from
early childhood to adolescence. Changes in proportions and body composition
during puberty are mirrored in somatotypes.
5. REFERENCES
AAHPER, 1965, Youth Fitness Test Manual. Washington D.C.
Bielicki, T., 1986, Physical growth as a measure of the economic well-being of population in twentieth
century. In Human Orowth (2nd ed.), edited by F. Falkner and l.M. Tanner (New York: Plenum
Press) Vol. 3 p. 283-305.
Bodzsar, B.E., 1982, The indices of the physique and the socio-economic factors based on a growth
study in Bakony girls. Anthropologiai KOzlemenyek (Budapest), 26, 129-134.
Bodzsar, B.E., 1991, The Bakony Orowth Study. Humanbiologia Budapestinensis, 22, pp. 210.
Carter, J.E.L., and Heath, H.B., 1990, Somatotyping. Development and Applications (Cambridge:
Cambridge University Press).
Duquet, W., 1980, Applicability of the Heath-Carter somatotype method to 6 to 13 year-old children [in
Flemish]. PhD Dissertation, Vrije Universiteit Brussel.
Eiben, 0.0., 1977, Changes in body measurements and proportions of children, based on the KOrmend
Orowth Study (in Hungarian with an English summary). Anthropologiai KOzlemenyek (Budapest),
21,43-58.
Eiben, 0.0., 1985, The KOrmend Orowth Study: Somatotypes. Humanbiologia Budapestinensis, 16,37-
52.
Eiben, 0.0., 1988a, History of Human Biology in Hungary. - Occasional Papers 2/4 pp 75. International
Association of Human Biologists, Newcastle upon Tyne, UK.
Eiben, 0.0., 1988b, Secular growth changes in Hungary. Secular trend of growth and biological
maturation of children in Hungary, based on the KOrmend Orowth Study (in Hungarian).
Humanbiologia Budapestinensis Suppl. 6., pp. 133.
Eiben, 0.0., and Panto, E., 1986, The Hungarian national growth standards. Anthropologiai
KOzlemenyek (Budapest), 30, 5-23.
Eiben, 0.0., and Panto, E., 1987/88, Body measurements in the Hungarian youth at the 1980s, based on
the Hungarian National Orowth Study. Anthropologiai KOzlemenyek (Budapest), 31,49-68.
Eiben, 0.0., Barabas, A., and Panto, E., 1991, The Hungarian National Orowth Study I. Reference data
on the biological developmental status and physical fitness of 3-18 year old Hungarian youth in the
1980s. Humanbiologia Budapestinensis, 1, pp. 123.
Farmosi, I., 1982, Results of constitutional and motor examinations of male students. Olasnik
AntropoloOkog Druotova Jugoslavije, 19,35-51.
Ouedes, D.P., 1983, Estudos antropometricos entre escolares. Revista Brazileire de Educacao Fisica e
Desportos, out/mar., 12-17.
Haag, H., and Dassel, H., 1975, Fitness Test. (Stuttgart: Verlag Karl Hofmann, Schorndorf bei
Stuttgart).
Heath, B.H., 1977, Applying the Heath-Carter somatotype method. - in Orowth and Development:
Physique edited by 0.0. Eiben (Budapest: Akademiai Kiado), p. 335-347.
Holopainen, S., Lumiaho, P., and Telarna, R., 1984, Level and rate of development of motor fitness,
motor abilities and skills by somatotype. Scandinavian Journal of Sport Sciences, 6/2, 67-75.
Kansal, D.K., 1981, A study of age changes in physique and body composition in males of two
communities in Punjab. PhD Thesis, Punjabi University, Patiala, India.
Martin, R, and Saller, K., 1957, Lehrbuch der Anthropologie, (Stuttgart: 1.-0. Fischer).
Matthews, D.R., 1973, Measurements in Physical Education. (Philadelphia: Saunders Co).
Nemeth, A. and Eiben, 0.0., 1997, Secular growth changes in Budapest in the 20th century. Acta
Medica Auxologica, 29, 5-12.
Nemeth, A., 1999, Somatic development of Budapest children and adolescents at the end of the XXth
Century and secular trend in Budapest. Ph.D. Thesis, Eotvos Lonind University, Budapest.
Nemeth, A., Bodzsar, B.E., and Eiben, 0.0., 1999, Comparison of fatness indicators in Budapest
children. Anthropologische Anzeiger, 57, 325-337.
Perez, B., Castillo, T.L., and Brief, F .K., 1985, Caracteristicas somatotipicas asociadas con la edad en
un grupo de escolares Venezolanos. Instituto de Investigaciones Economicas y Sociales. FACES
Universidad Central de Venezuela, Caracas, Venezuela.
Rangan, S.C.B., 1982, Validity of age, socio-economic belonging and dietary type as somatotype
determinants in boys of secondary schools. PhD Thesis, Bangalore University, Bangalore, India.
Simon, 1. ,and Renson, R., 1982, Evaluation of Motor Fitness. Report of the European Seminar on the
Evaluation of Motor Fitness. Leuven University Press, Leuven.
308 CHAPTER 23
Singh, S.P., and Sidhu, L.S., 1980, Changes in somatotypes during 4 to 20 years in Gaddi Rajput boys.
Zeitschrift fur Morphologie und Anthropologie, 71, 285-293.
Slaughter, M., Lohman, T.G., and Misner, J.E., 1980, Association of somatotype and body composition
to physical performance in 7-12 year old girls. Journal of Sports Medicine and Physical Fitness, 20,
189-198.
Stepnicka, J., 1976, Somatotypes of Bohemian and Moravian youth. acta Facultatis Medicinae
Universitatis Brunensis, Brno, 57,233-242.
Szmodis, I., 1977, Physique and growth estimated by Conrad's and Heath-Carter's somatocharts in
athletic children. In Growth and Development: Physique edited by O.G. Eiben (Budapest:
Akademiai Kiad6), p. 407-415.
Tanner, J.M, Hiernaux, J., and Jarman, S., 1969, Growth and physique studies. In Human Biology. A
Guide to Field Methods: IBP Handbook 9 edited by J.S. Weiner, J.A. and Lourie
(OxfordlEdinburgh: Blackwell Scientific Publishers) p. 1-76.
Tanner, J.M., and Whitehouse, R.H., 1982, Atlas of Children's Growth. Normal Variation and Growth
Disorders (New York: Academic Press).
Tanner, JM., 1978, Human growth standards: Contribution and use. In Auxology: Human Growth in
Health and Disorders, edited by I. Gedda, and P. Parisi (London: Academic Press), p. 109-122.
Toriola, A.L., and Igbokwe, N.U., 1985. Relationship between perceived physique and somatotype
characteristics of 10 to 18 year old boys and girls. Perceptual and Motor Skills, 60, 878.
WHO, 1976, New Trends and Approaches in the Delivery of Maternal and Child Care in Health
Services. Sixth report of the WHO Expert Committee on Maternal and Child Care, WHO Technical
Report Series no 600. World Health Organisation, Geneva.
SOMATOTYPES OF BUDAPEST CHILDREN 309
APPENDIX
M. PROKOPEC
National Institute of Public Health, Prague, Czech Republic
l. INTRODUCTION
Growth of the human body from birth to adulthood does not proceed as a simple
enlargement of the body tissues, organs and parts. It is obvious that body
proportions change considerably from the new-born to the grown-up male or female.
If we take the sizes of individual body parts of a child at the age of one month as a
starting point (zero) and the sizes of the same body parts at the age of 20 years as
final (100 percent), we get different rates of growth of individual body measurements
at each chosen age.
The body of the growing child and youth changes its proportions year by year.
Lengths, widths, circumferences of the body and of its parts and body weight set out
for their individual journeys from one month of age towards their goal in early
adulthood at the age of 20. They proceed at different speeds reflecting thus the
changing morphology and individuality of the child in the course of its pre-school
and school years, and throughout its teens until each of them reaches its 100 percent
fInal size at the age of 20 where they all meet (Marshall 1977).
Ages at which the CUIVes of individual measurements are most wide apart in their
paths may indicate a period of unbalanced, unconsolidated, disproportionate state of
the body which may be associated with greater demand on energy regulating body
integrity and thus resulting in an increased health risk. It may be useful to know at
which age this time period comes in a child population as a whole or in an
individual. The fIrst decade of life is a period at which the quickly developed head
(and brain) governs the yet much less developed body parts which may be an
advantage in learning.
The body proportionality of the human being which has been reached at the age
of 20 developed hundreds of thousands of years and may be considered as stabilised,
as a fInal product of human evolution and any stage preceding it may be thus viewed
as immature, unfinished, as a state of transition. Naturally, some individuals may
reach the mature stage earlier than at 20 years, say already at the age of 16 in females
and 18 in males. It is obvious that further changes in the human body take place
after the age of 20 years, but they need not be considered here as important. On the
313
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 313-320.
© 2001 Kluwer Academic Publishers.
314 CHAPTER 24
other hand the body proportions of a physiologically normal new-born are the
product of its prenatal development "in utero" and must be taken as granted.
The aim of this paper is:
• To show the widely different rates of growth of various body parts
(measurements) in boys and girls followed-up from one month to 20 years of
age.
• To study changes in body proportions throughout the entire growing period,
and to enable to determine specific body proportions (mutual relations of
various body parts) at any given age.
• To point out ages at which the differences in the rate of growth in individual
measurements are greatest and when they seem unimportant.
• To detect mutual similarities and diversities in the pattern of growth between
individual parts (measurements) of the body.
• To determine differences in boys and girls in this respect.
3. RESULTS
The results are given in Table 1 for both sexes and in Figure 1 for males and in
Figure 2 for females.
The following measurements: bicristal diameter, stature, sitting height and
biacromial diameter had similar patterns of growth. Thus, the curve of the bicristal
diameter represents the rest of these measurements in the figures. The type of the
curve of the chest circumference is similar to upper arm circumference. The latter is
thus not shown in the figures.
An extremely wide variability in the rate of growth of individual measurements
appear from Table 1 and from Figures 1 and 2. The majority of curves follow the
typical trend of stature (a steep start from one month to 3 years, followed by smaller
and regular increments until pubertal spurt after which a transition to a steady state
follows). This is not the case in head measurements (head breadth, head length and
head circumference) the CUIVes of which are most advanced of all others straight from
the beginning. Their CUIVes are followed by those of the femur and humerus, which
intersect the former CUIVes at the age from 12 to 14 years in boys. Only the CUIVe of
the humerus does the same between 9 and 12 years in girls.
The CUIVe of the bicristal diameter (which represents also stature, sitting height
and biacromial diameter) occupies the medium position among other growth CUIVes
in both sexes. The curve of the calf intersects that of the bicristal width at the age of
about 10 years both in boys and girls. The calf CUIVe in boys runs alone before the
age often, whereas in girls it twists with that of abdomen circumference in the same
period of time. The curves of abdominal and chest circumferences follow the same
path until the age of 13 in boys, whereas in girls they run separately but cross before
the age of 11 years.
The CUIVe of weight runs wide apart from the curves of other body parts in both
sexes, showing thus how much weight lagged behind the other measurements. It
remains separate until the age of 20 in girls, but it has been crossed by the CUIVe of
chest circumference shortly after 14 years of age in boys.
316 CHAPTER 24
100
90 L head br.
head c.
80
~ 60
~
Q.. 50
"-humerus
40 biiliocrist. d.
30
20 Boys
10
0
0 2 4 6 8 10 12 14 16 18 20
Age, years
\00
90
80
70
....
5u 60
[)
Q.. 50
40 biiliocrist. d.
30
20 Girls
10
0
0 2 4 6 8 10 12 \4 16 18 20
Age, years
Table 1. Age (in years) at which children attain the given percent of total increment
from 1 month to 20 years of age (M: male; F: female)
Percent 10 20 30 40 50
M F M F M F M F M F
Weight 1.0 0.8 4.0 3.3 6.8 5.9 9.2 8.0 11.3 9.5
Stature 0.5 0.6 1.3 1.2 2.5 2.2 4.0 3.5 6.0 5.3
Sitting h. 0.3 OA 0.8 0.9 2.0 1.8 3.6 3.2 6.0 5.3
Chest eire. 0.3 0.3 0.8 0.7 3.3 2.5 6.1 5.6 8.5 7.6
Head eire. 0.1 0.1 0.2 0.2 OA 0.5 0.7 0.6 1.2 1.1
Upp. arm e. 0.3 0.2 0.5 0.5 2.5 lA 6A 5.0 9.0 7.3
Calf eire. 0.2 0.2 0.5 0.6 1.6 l.6 4.8 4.1 7.0 6.6
Femur 0.2 0.3 OA 0.5 l.0 0.9 2.9 1.7 3A 3.0
Humerus 0.3 0.3 0.7 0.5 1.5 l.0 3.2 2.1 5.2 3.6
Bieristal w. 0.2 0.3 0.8 0.8 1.8 1.6 3.2 3.0 5.2 5.1
Biaerom. w. 0.3 OA 1.1 l.0 2.3 l.9 4.2 3A 6.2 5.1
Abdom. eire. 0.2 0.2 0.7 0.6 3.5 l.6 6.0 4.2 8.3 6.5
Head breadth 0.1 0.1 0.3 0.2 OA 0.3 0.6 0.6 0.9 0.8
Head length 0.2 0.2 OA OA 0.6 0.6 0.9 0.8 1.5 1.3
Percent 60 70 80 90 100
M F M F M F M F M F
Weight 12.6 10.9 13.7 12.0 15.0 13.2 l6A 15.0 20.0 20.0
Stature 8.0 7.1 lOA 9.0 12.7 10.9 14.3 12.6 19.0 18.0
Sitting h. 8.6 7.5 11.5 9.7 13.3 11.7 15.0 13.3 20.0 19.0
Chest eire. 10.8 9.3 12.9 10.7 14.3 12.2 16.0 13.9 20.0 19.0
Head eire. 2A 2.0 5.3 4.5 9.0 7.2 14.0 11.5 20.0 20.0
Upp. arm e. 11.0 9.0 12.9 10.7 14.6 12.9 16.3 14.6 20.0 18.6
Calf eire. 9.1 8.6 10.6 lOA 12A 11.9 14A 13.8 19.0 18.0
Femur 5.5 4.7 7.9 7.0 10.8 9.0 13.3 11.8 19.0 19.0
Humerus 7.3 5.1 9.7 6.7 12A 8.6 13.9 10.7 19.0 19.0
Bieristal w. 8.0 7.6 10.7 9.8 12.8 11.8 14.7 13.7 19.0 19.0
Biaerom. w. 8.6 7.2 11A 9.2 13A 11.2 15.0 13.2 20.0 19.0
Abdom. eire. 10.5 8.5 12.9 10.6 14.9 12.7 16.8 14.7 20.0 19.9
Head breadth 1.6 1.3 3.2 2.2 6.0 4.1 13.3 8.2 20.0 20.0
Head length 2.7 2.2 5.1 3.9 9.7 6.8 15.0 11.3 20.0 20.0
The greatest difference in years between the curves of the head breadth and weight
(extremes), was found to be in boys at the level of 60 percent and amounts to 12
years of time space. The same was found in girls at the level of 70 percent and
amounts to 10.5 years of time space.
The most rapid rate of growth show the head measurements. The head breadth
reached the level of 80 percent of the total increment from 1 month to 20 years at the
age of 6 years in boys and at the age of 4.1 years in girls. The head circumference
and the head length reached the level of 80 percent in boys at 9 and 9.7 years,
respectively; and in girls at 7.2 and at 6.8 years, respectively.
318 CHAPTER 24
The post-pubertal spurt in head measurements (only in boys after the age of 14
years) may be explained by the development of frontal sinuses and by thickening of
the skull bones and skin at that age. This is a visibly clear difference between boys
and girls in Figures 1 and 2.
The following picture may be seen in boys at the level of 50 percent of the total
increment from 1 month to 20 years of age:
The head breadth reached this level at 0.9 years
head circumference at 1.2 years
head length at 1.5 years
biepicondylar femur at 3.4 years
biepicondylar humerus and bicristal diameter at 5.2 years
stature and sitting height at 6.0 years
biacromial diameter at 6.2 years
calf circumferen::e at 7.0 years
chest circumference at 8,5 years
upper ann circmnference at 9.0 years
abdominal circmnferences at 8.3 years
and weight at 11.3 years
Similarly, ages may be easily read from Table 1 and/or from Figures 1 and 2 for
any body dimension at the aibitraIy percent level in boys and in girls.
Different rates of growth in the measurements followed up in this study may be
also detected from Figures 1 an 2 when we look at aibitraIy percent levels, reached
by individual measurements at given ages. The greatest span is between the curves
of head breadth and body weight. In boys at the ages from 3 to 6 years it amounts to
53 percent of the total increment from 1 month to 20 years. Whereas weight attained
at the age of three years represents only 17 percent of the total increment, head
breadth attained at the same age equals 71 percent of the total increment.
At 6 years the body weight reaches in boys 27 percent of the total increment,
circumferences of the thorax (also upper arm) and abdomen reach 40 percent; calf
circumference 45 percent; bicristal diameter (and also stature, sitting height and
biacromial diameter), 54 percent; biepicondylar humerus, 55 percent; biepicondylar
PREPARATION OF A CAMERA-READY MANUSCRIPT 319
femur 63 percent; head circumference and head length, 73 percent; and head breadth,
81 percent of the total increment from 1 month to 20 years.
The greatest span between weight curve and head breadth curve in girls is at the
age of 4 years. It amounts to 56 percent of the total increment. At this age the
weight reaches only 22 percent and head breadth, 79 percent of the total increment.
Weight reached in girls at the age of 6 years, 30 percent of the total increment;
thorax circumference (and the upper arm circumference), 42 percent; abdominal and
calf circumferences, 47 percent; bicristal (and also stature, sitting height and
biacromial diameter), 53 percent; biepicondylar humerus, 65 and biepicondylar
femur, 66 percent; head circumference and head length, 76 and 77 percent,
respectively; and head breadth, 86 percent.
4. DISCUSSION
We may postulate, that ideal proportions of the human body are attained at early
adulthood (say, 18 to 20 years), which is the result of hundreds of thousands years
long development of man on the Earth. Thus we may consider human proportions
between birth and 20 years as tmnsitional, unbalanced, unfinished, and suspect the
time period of greatest diversity in the rate of grow of the body parts to be a health
risk. On the other hand, we cannot deny beauty in children at any given age, or
body efficiency in sports to teenagers with still evident disproportions between trunk
and extremities.
We may also deduce from the amazingly great variability in the rate of growth of
individual body parts that the development of an individual works stepwise,
concentrating on one structure at a time, and on another one when its time comes.
Throughout the period of child development from birth to the teens the body is
dominated by the growth of the head. The necessary body structures (skeleton,
muscles of the extremities) develop less quickly, later the chest and trunk develop,
but weight is minimal and the capacity of neurones in the brain may serve learning
much more effectively than later in life. Starting at puberty the body comes to power
and the reign of the brain is diminishing.
There is much beauty in the unfinished human with a big head and big eyes.
The will to learn and the dependence on society makes the young preferred, beloved
and well treated. Their relative rareness in the society helps to support such an
attitude (children under 14 years make up for a fraction of the population in an
industrialised country). Once one enters the domain of adults, one is lost in
triviality. There are too many adults in an ageing society.
After all, the period of life during which the man and woman are in the making is
miraculous.
5. CONCLUSION
The ages at which boys and girls attain a certain percentage of the total increment
from 1 month to 20 years of age showed that girls are ahead of boys in all
measurements. At the level of 50 percent of the total increment from 1 month to 20
years of age the difference between the sexes amounts to 0.1 years in the head
320 CHAPTER 24
Acknowledgement. The author thanks Marcela Havlinova for drawing the figures
and Dr. Pavel Chyle for the revision of the English text.
6. REFERENCES
Falkner, F., 1955, A baseline of investigations for longitudinal growth studies in the child. Centre
International de l'Enfance, Chateau de Longchamp, Bois de Boulogne, Paris XVIo.
Kapalin, V., Kotaskova, J., and Prokopec, M., 1969, Physical and Mental Development of the
Contemporary Generation of Our Children. [In Czech, English Summary] (Prague: Academia) 300
pp.
Marshall, W.A., 1977, Human Growth and its Disorders (London, New York, San Francisco: Academic
Press) p.179.
Prokopec, M., 1980, Longitudinal Follow-up of Physical and Mental Development and Health State of a
Selected Group of Children from Prague: Final Report [In Czech] (Prague: National Institute of
Public Health) 126 pp.
Tanner, 1M., 1947, A Guide to American Growth Studies. Yearbook of Physical Anthropology, 3, 28-
33.
CHAPTER 25
SHORT-TERM GROWTH
M. HERMANUSSEN
1. INTRODUCTION
to 4-fold variation in the intra- and inter-individual monthly lower leg growth rates,
and they failed to provide evidence for any correlation between one month lower leg
growth and annual height growth, when analysing short-term measurements of lower
leg length and body stature in 90 healthy children aged 3-16 years. Several authors
concluded that growth of different parts of the skeleton and variable interval growth
rates limit the ability of knemometry to predict long-term growth (literature reviewed
in Hermanussen 1998).
It is questionable whether short-term length increments should be used to predict
long-term growth, at all. From the very beginning, several authors have
demonstrated significant non-linearity of short-term growth by the use of weekly
measurements of lower leg length, and though most authors did not further
characterise the dynamics of short-term growth, they have made clear that linear
extrapolations of non-linear events do not appear to be very useful.
the observed distributions with those predicted by the saltatory, and by continuous
growth models. The authors concluded that their data were incompatible with the
saltatory model. Similar findings were published by us, and further longitudinal
studies have shown that only a minority of gain in stature can be said to be truly
saltatory. It was suggested that alternate series of saltatory events and stasis simply
are a result of the assumptions of the mathematical model, as the model only allows
"increase" (saltation), or "no increase" (stasis). The model does not seem applicable
for truly linear growth (Herman us sen 1998).
has been named an "incremental burst" or "mini growth spurt". These local
structures were represented by parameterised functions, in this case, by so-called
Gompertz functions, i.e. double exponential functions that contain a central S-shaped
structure.
Series of measurements were converted into series of time windows, containing
either exactly one Gompertz function or no Gompertz function. If a Gompertz
function was detected within a time window, the respective "mini growth spurt"
could be characterised by the three parameters of this function, i.e., amplitude,
inflection point (the day of peak growth velocity), and the slope. The slope
determines rapidity and, thus, alludes to the duration that one incremental burst needs
for completion. In healthy neonates (Herman us sen 1998), we found a mean
amplitude of mini growth spurts of 2057 ~m (SD 1132 ~m) and a mean peak
growth velocity of 85 ~m per hour. The spurts appeared once every 4.2 days on
average, but the standard deviation of 1.9 days indicated that the time interval
between subsequent spurts varied considerably. Time series analysis confirmed that
mini growth spurts did not occur periodically.
Similar characteristics of mini growth spurts were detected in growing rats. 5018
quadruplet daily measurements of lower leg length were performed in 62 female, and
81 male rats. We detected and analysed 684 local structures (mini growth spurts) and
found a mean amplitude of 2153 ~m (SD 1034 ~m) in female rats, and of 2958 ~m
(SD 1614 ~m) in males, with peak growth velocities that were lower than in
humans. Partial growth hormone deficiency led to a modification of mini growth
spurts in the rats with a reduction of spurt amplitude, and an increase in peak growth
velocity that was both reversed when exogenous growth hormone was administered.
The interval between subsequent mini growth spurts ranged between 4.2 and 4.6
days, with a large standard deviation (Hermanussen et at. 1998).
First in animal studies and later in neonates (Keller et af. 1998), we detected
growth periods that contained no mini growth spurts. The lower leg length increased
linearly, but the respective time windows appeared "empty", i.e., the algorithm was
unable to detect an S-shaped structure of the increments. Thus, evidence had occurred
that short-term growth may not always proceed in spurts. This must be taken into
consideration when applying mathematical models for the description of short-term
growth.
increments, but it remained uncertain whether this effect was a brief soft tissue
reaction similar to that seen after growth hormone administration (Hermanussen,
Sippell 1985), or whether it presented the beginning of a long-term stimulation of
growth. In order to study the dynamics of short-term growth, we investigated the
growth pattern that followed the transfer from three times weekly intra-muscular to
daily sub-cutaneous administration of human growth hormone in growth hormone
deficient patients (literature reviewed in Hermanussen 1998). Though the
measurements were not done at strict weekly intervals, a significant catch-up growth
spurt was detected immediately after transfer of administration which lasted for
several weeks and was followed by an intermittent decrease of growth velocity.
Thereafter, growth velocity proceeded in a wave-like pattern, and stabilised on a
significantly higher level than before the transfer of application. Later studies of
short-term growth that were based on strict weekly series of lower leg measurements
have confirmed this pattern (Hermanussen 1998). It appeared logical to use short-
term growth as predictive criterion for successful growth promotion in growth
hormone therapy. Serial measurements of the lower leg during the first 10 weeks of
growth hormone administration provided useful information in 21 out of 24 children
on the growth response during the first year of treatment. Weekly measurement
regimens were used for optimising growth therapies (Hermanussen 1995), and we
investigated the possibility to synchronise the administration of growth hormone
with the spontaneous pattern of short-term growth (Hermanussen 1998).
Other authors used weekly measurements of the lower leg length for studying
growth promotion during treatments with growth hormone (Guzzaloni et al. 1997a,
Wales, Milner, 1989, Wit et al. 1987). But most of these studies lacked adequate
statistical treatment of the weekly measurement series. In view of the traditional
ideas of short-term growth being a linear process, growth was usually expressed in
terms of "linear growth rates", i.e. by length differences within intervals of 3 to 6
weeks. And since serial measurements at weekly intervals are tedious, many
investigators avoided the weekly series, and only obtained measurements once every
3 to 6 weeks, with no measurements in-between. In other words, many "short-term"
growth studies consist of subsequent 3 to 6-week length increments.
Ross et at. (1986) used four-week increments in children with Turner syndrome,
and compared mean leg length increment before and during treatment with growth
hormone, similar protocols were used by Gelato et al. (1985) who studied the effects
of pulsatile administration of growth hormone-releasing hormone in children with
growth hormone deficiency. Mortensen et at. (1991) published that lower leg growth
during the first three months of growth hormone treatment was significantly
associated with total height gain during the following nine months period, and when
comparing 4-week periods of growth hormone treatment in patients with Turner
Syndrome, Wales and Milner 1987b identified those children in whom growth
hormone treatment was not beneficial (literature reviewed in Hermanussen 1998).
When pre-treatment "linear growth rates" remained significantly smaller than "linear
growth rates" during treatment, it was assumed that growth stimulation had taken
place. Knemometry has also been applied in patients with constitutional growth
delay who were treated with oxandrolone (Guzzaloni et at. 1997b).
Since knemometry is not available for everybody, studies have started to replace
measurement accuracy by measurement frequency. In view of the brilliant
descriptions of growth by daily series of body height measurements, we inaugurated
328 CHAPTER 25
11. REFERENCES
Agertoft, L., and Pedersen, S., 1997, Short-term knemometry and urine cortisol excretion in children
treated with fluticasone propionate and budesonide: a dose response study. European Respiratory
Journal, 10, 1507-12.
Agertoft, L., and Pedersen, S., 1999, Short-term lower leg growth rate in children with rhinitis treated
with intranasal mometasone furoate and budesonide. Journal of Allergy and Clinical Immunology,
104, 948-52.
Ahmed, S.F., Wallace, W.H., Crofton, P.M., Wardhaugh, B., Magowa, R., and Kelnar, C.J., 1999,
Short-term changes in lower leg length in children treated for acute lymphoblastic leukaemia.
Journal of Pediatric Endocrinology & Metabolism, 12, 75-80.
Ahmed, S.F., Wallace, W.H.B., and Kelnar, C.J.H., 1995, Knemometry in childhood: a study to
compare the precision of two different techniques. Annals of Human Biology, 22, 247-252.
Ahmed, S.F., Wardhaugh, B.W., Duff, J., Wallace, W.H.B., and Kelnar, C.J.H., 1996, The relationship
between short-term changes in weight and lower leg length in children and young adults. Annals of
Human Biology, 23,159-162.
Ashizawa, K., and Kawabata, M., 1990, Daily measurements of the heights of two children from June
1984 to May 1985. Annals of Human Biology, 17,437-443.
330 CHAPTER 25
Boyd, E., 1980, Origins of the study of human growth. Eugene OR, University of Oregon Health
Science Foundation.
Camerer, W., 1880, Versuche tiber den Stoffwechsel, angestelit mit 5 Kindem im Alter von 2 bis 11
Jahren. Zeitschrift ftir Biologie, 16, 24-41.
Cianfarani, S., Germani, D., Rossi, P., Rossi, L., Germani, A., Ossicini, c., Zuppa, A., Argiro, G., Holly,
J.M., and Branca, F., 1998, Intrauterine growth retardation: evidence for the activation of the
insulin-like growth factor (IGF)-related growth-promoting machinery and the presence of a cation-
independent IGF binding protein-3 proteolytic activity by two months of life. Pediatric Research,
44,374-380.
Crofton, P.M., Ahmed, S.F., Wade, J.C., Elmlinger, M.W., Ranke M.B., Kelnar C.J., and Wallace W.,
1999, Effects of a third intensification block of chemotherapy on bone and collagen turnover,
insulin-like growth factor I, its binding proteins and short-term growth in children with acute
lymphoblastic leukaemia. European Journal of Cancer, 35, 960-967.
Crofton, P.M., Ahmed, S.F., Wade, J.C., Elmlinger, M.W., Ranke M.B., Kelnar, C.J., and Wallace, W.,
2000, Bone turnover and growth during and after continuing chemotherapy in children with acute
lymphoblastic leukemia. Pediatric Research, 48, 490-496.
Crofton, P.M., Shrivastava, A., Wade, J.C., Stephen, R., Kelnar, C.J.H., Lyon, A.J., and Mcintosh, N.,
1996, Bone and collagen markers in preterm infants: relationship with growth and bone mineral
content over the first 10 weeks of life. Pediatric Research, 46, 581-587.
Cronk, C.E., Stallings, V.A., Spender, Q.W., Ross, J.L., and Widdoes, H.D., 1989, Measurement of short
term growth with a new knee height measuring device (KHMD). American Journal of Human
Biology, 1,421-428.
Davies, H.A., Pickering, M., and Wales, J.K.H., 1996, A portable knemometer: a technique for
assessment of short-term growth. Annals of Human Biology, 23, 149-157.
Gelato, M.C., Ross, J.L., Malozowski, S., Pescovitz, O.H., Skerda, M., Cassoria, F., Loriaux, L., and
Merriam, G.R., 1985, Effects of pulsatile administration of growth hormone (GH)-releasing
hormone on short term linear growth in children with GH deficiency. Journal of Clinical
Endocrinology and Metabolism, 61, 444-450.
Gibson, A.T., Pearse, R.G., and Wales, J.K.H., 1993, Knemometry and the assessment of growth in
premature babies. Archives of Disease in Childhood, 69, 498-504.
Glock, M., Hermanussen, M., Keller, E., and Hartmann, K.K.P., 1999, Gulliver G-100 - a new device
to evaluate daily growth measurement in comparison with Harpenden stadiometer. Hormone
Research, 52, 287-290.
Guzzaloni, G., Grugni, G., Moro, D., Minocci, A., and Morabito, F., 1997a, Impiego della knemometria
nella valutazione degli effetti a breve termine della terapia con GH in bambini con diversi disordini
accrescitivi. Minerva Pediatrica, 49, 139-145.
Guzzaloni, G., Grugni, G., Moro, D., Minocci, A., and Morabito, F., 1997b, Usefullness of knemometry
in the assessment of oxandrolone effects and role of infections in short-term growth within
constitutional growth delay. Acta Medica Auxologica, 29, 21-27.
Heinrichs, c., Munson, P.J., Counts, D.R., Cutler, G.B., and Baron, J., 1995, Patterns of human growth.
Science, 268, 442-444.
Hermanussen, M., 1995, Growth promotion by steroid substitution in short stature treated with GH and
LHRH analogue. Acta Paediatrica, 84, 1207-1210.
Hermanussen, M., 1998, The analysis of short-term growth. Hormone Research, 49, 53-64.
Hermanussen, M., and Seele, K., 1997, Mini-knemometry: an accurate technique for lower leg length
measurements in early childhood. Annals of Human Biology, 24, 307-313.
Hermanussen, M., and Sippell, W.G., 1985, Knemometric monitoring of early effects of human growth
hormone on leg growth in children with growth hormone deficiency, Lancet, I, 1069-1071.
Hermanussen, M., Bugiel, S., Aronson, S., and Moell, C., 1992, A non-invasive technique for the
accurate measurement of leg length in animals. Growth, Development & Aging, 56,129-140.
Hermanussen, M., Geiger-Benoit, K., and Sippell, W.G., 1986, GH deficient children receiving GH
replacement do not grow during intermittent infectious illness. Acta Paediatrica Scandinavica, 75,
601-604.
Hermanussen, M., Geiger-Benoit, K., Burmeister, J., and Sippell, W.G., 1988a, Knemometry in
childhood: accuracy and standardization of a new technique of lower leg length measurement.
Annals of Human Biology, 15, 1-16.
Hermanussen, M., Geiger-Benoit, K., Burmeister, J., and Sippell, W.G., 1988b, Periodical changes of
short term growth velocity ("mini growth spurts") in human growth. Annals of Hurnan Biology, 15,
103-109.
Hermanussen, M., Rol de Lama, M., Burmeister, J., and Fernandez-Tresguerres, J.A., 1995, Mikro-
knemometry: an accurate technique of growth measurement in rats. Physiology nd Behavior, 58,
347-352.
Hermanussen, M., Rol de Lama, M., Perez Romero, A., Ariznaverreta, C., Burmeister, J., and
Tresguerres, J.A.F., 1996, Differential catch-up in body weight and bone growth after short term
starvation in rats. Growth Regulation, 6, 230-237.
SHORT-TERM GROWTH 331
Hermanussen, M., Rol de Lama, M., Tresguerres, J.A.F., Grasedyck, L., and Burmeister, J., 1998, Short
term growth: evidence for chaotic series of mini growth spurts in rat growth. Physiology and
Behavior, 64, 7-13.
Heuck, C., and Wolthers, 0.0., 1997, Calculation of knemometric growth rates in group studies of
children treated with exogenous glucocorticoids. Annals of Human Biology, 24, 411-418.
Heuck, c., Heickendorff, L., and Wolthers, 0.0., 2000, A randomised controlled trial of short term
growth and collagen turnover in asthmatics treated with inhaled formoterol and budesonide.
Archives of Disease in Childhood, 83, 334-339.
Heuck, c., Ternowitz, T., Herlin, T., and Wolthers, 0.0., 1998a, Knemometry in children with atopic
dermatitis treated with topical glucocorticoids. Pediatric Dermatology, 15,7-11.
Heuck, C., Wolthers, 0.0., and Herlin, T., 1999, growth-suppressive effect of intra-articular
glucocorticoids detected by knemometry. Hormone Research, 52, 95-Q6.
Heuck, c., Wolthers, 0.0., Kollerup, G., Hansen, M., and Teisner, B., 1998b, Adverse effects of
inhaled budesonide (880f.lg) on growth and collagen turnover in children with asthma: a double-
blind comparison of once-daily versus twice-daily administration. Journal of Pediatrics, 133, 608-
612.
Hulanicka, B., Gronkiewicz, L., and Koziel, S., 1999, Wzrastanie dzieci badanie knemometryczne.
Monografie. Zakladu Antropologii Polskiej Akademii Nauk. Wroclaw.
Kaempf, D.E., Pfliiger, M.S., Thiele, A.M., Hermanussen, M., and Linderkamp, 0., 1998, Influence of
nutrition on growth in premature infants: assessment by knemometry. Annals of Human Biology, 25,
127-136.
Keller, A., Hermanussen, M., Vogtmann, C., Kiess, W., and Keller, E., 1999, Effect of erythrocyte
transfusion on longitudinal bone growth of premature infants assessed by mini-knemometry.
European Journal of Pediatrics, 158,871-872.
Keller, A., Hermanussen, M., Vogtmann, Ch., Burmeister, 1., and Keller, E., 2000a, Analysis of short-
term growth of premature infants during the first days of life assessed by mini-knemometry. Acta
Medica Auxologicaogica, 32, 40.
Keller, A., Keller, E., Burmeister, J., and Hermanussen, M., 1998, Daily knemometric measurements
can visualize the clinical situation of critically ill infants. Hormone Research 50, suppl 3, 37.
Keller, A., Vogtmann, Ch., Hermanussen, M., and Keller, E., 2000b, Short-term growth during the first
month of life of SGA and AGA infants. In: II. Aschauer Seminar: The impact of environmental
factors on individual development with special reference to nutrition and obesity. Monatschrift fiir
Kinderheilkunde, 148, 810-815.
Keller, A., Vogtmann, Ch., Keller, E., Hermanussen, M., and Kiess, W., 2000c, Differences of short-
term growth in relation to varying doses of dexamethasone administration in premature infants.
Hormone Research 53 (suppl 2), 29
Klein, K.O., Munson, P.J., Bacher, J.D., Cutler, G.B., and Baron, J., 1994, Linear growth in the rabbit is
continuous, not saltatory. Endocrinology, 134, 1317-1320.
Kobayashi, M., and Togo, M., 1993, Twice-daily measurements of stature and body weight in two
children and one adult. American Journal of Human Biology, 5, 193-201.
Lampl, M., and Johnson, M.L., 2000, Distribution methods and analysis of nonlinear longitudinal data.
Methods in Enzymology, 321,182-195.
Lampl, M., Ashizawa, K., Kawabata, M., and Johnson, M.L., 1998, An example of variation and
pattern in saltation and stasis growth dynamics. Annals of Human Biology, 25, 203-219.
Lampl, M., Veldhuis, J.D., and Johnson, M.L., 1992, Saltation and stasis: a model of human growth.
Science, 258, 801-803.
Ledford, A.W., and Cole, T.J., 1998, Mathematical models of growth in stature throughout childhood.
Annals of Human Biology, 25,101-115.
MaIling-Hansen, R., 1886, Periods in the growth of children and in the heat of the sun. Copenhagen,
Vilhelm Tryde; cited by: Michaelsen KM, 1994, Short-term measurements of linear growth using
knemometry. Journal of Pediatric Endocrinology, 7, 147-154.
Marshall, W.A., 1971, Evaluation of growth rate in height over periods of less than one year. Archives
of Disease in Childhood, 46, 414-420.
Michaelsen, K.F., Skov, L., Badsberg, J.H., and Jorgensen, M., 1991, Short-term measurement of linear
growth in preterm infants: validation of a hand-held knemometer. Pediatric Research, 30, 464-468.
Milani, S., 2000, Kinetic models for normal and impaired growth. Annals of Human Biology, 27,1-18.
Mortensen, H.B., Main, K., Michaelsen, K.F., Kastrup, K.W., Jorgensen, J.T., and Skakkebaek, N.E.,
1991, Predicting and monitoring of growth in children with short stature during the first year of
growth treatment. Acta Paediatrica Scandinavica, 80, 1150-1157.
Nicholl, R.M., and Gamsu, H.R., 1999, Changes in growth and metabolism in very low birth weight
infants fed with fortified breast milk. Acta Paediatrica, 88, 1056-1O6\.
Rol de Lama, M.A., Perez-Romero, A., Hermanussen, M., Ariznavarreta, R.C., and Tresguerres,
J.A.F., 1998, Sexual dimorphism in growth as measured by microknemometry: Different responses
to GH deficiency and exogenous GH administration. Neuroendocrinology, 68, 210-219.
332 CHAPTER 25
Rol de Lama, M.A., Perez-Romero, A., Tresguerres, 1.A.F., Hermanussen, M., and Ariznavarreta,
R.C., 2000, Recombinant human growth hormone, rhGH, enhances tibial growth in peripubertal
female rats but not in males. European lournal of Endocrinology, 142,517-523.
Ross, 1.L., Long, L.M., Skerda, M., Cassorla, F.G., Loriaux, D.L., and Cutler, G.B., 1986, Growth
response relationship between growth hormone dose and short term growth in patients with Turner's
syndrome. lournal of Clinical Endocrinology and Metabolism, 63, 1028-1030.
Seidel, Chr., Schafer, F., Walther, U., and Schlirer, K., 1991, The application of knemometry in renal
disease: preliminary observations. Pediatric Nephrology, 5, 467-471.
Shrivastava, A., Lyon, A., and Mcintosh, N., 2000, The effect of dexamethasone on growth. Mineral
balance and bone mineralisation in preterm infants with chronic lung disease. European 10urnal of
Pediatrics, 159, 380-384.
Tanner, 1.M., 1978, Foetus into man. 1st ed. London, Open Books Publishing Ltd, pp 7-8.
Togo, M., and Togo, T., 1982, Time-series analysis of stature and body weight in five siblings. Annals
of Human Biology, 9, 425-440.
Valk, I.M., 1971, Accurate measurement of the length of the ulna and its application in growth
measurements. Growth, 35, 297-310.
Valk, I.M., Langhout Chabloz, A.M.E., and van Gilst, W., 1983b, Intradaily variation of the human
lower leg length and short term growth: a longitudinal study in fourteen children. Growth, 47, 397-
402.
Valk, I.M., Langhout Chabloz, A.M.E., Smals, A.G.H., Kloppenborg, P.W.C., Cassorla, F.G., and
Schutte, E.A.S.T., 1983a, Accurate measurement of the lower leg length and the ulnar length and
its application in short term growth measurements. Growth, 47, 53-66.
Van den Bosch, 1.S.G., Smals, A.G.H., Pieters, G.F.F., Valk, I.M., and Kloppenborg, P.W.c., 1981, The
effect of low dose oestrogens on short term growth and concomitant biochemical phenomena in
girls with tall stature. Acta Endocrinologica, 98, 156.
Visser, MJ., van Aalderen W.M., Elliott B.M., Odink RJ., and Brand P.L., 1998, Short-term growth in
asthmatic children using fluticasone propionate. Chest, 113,584-586.
Wales, 1.K.H., and Gibson, A.T., 1994, Short-term growth - rhythms, chaos or noise? Archives of
Disease in Childhood 71,84-89.
Wales, 1.K.H., and Milner, R.D.G., 1987a, Knemometry in the assessment of linear growth. Archives of
Disease in Childhood, 62, 166-171.
Wales, J.K.H., and Milner, R.D.G., 1987b, Knemometry as a predictor of response to somatrem in
Turner's syndrome. Acta Paediatrica Scandinavica Supp!., 337, 37-39.
Wales, 1.K.H., and Milner, R.D.G., 1989, Changes in leg length and height during treatment with
somatropin. Archives of Disease in Childhood, 64,1708-1712.
Wit, 1.M., van Kalsbeek, EJ,. von Wijk-Hoek I.M., and Leppink, G.1., 1987, Assessment of the
usefulness of weekly knemometric measurements in growth studies. Acta Paediatrica
Scandinavica, 76, 974-980.
Wolthers, O.D., 1996, Long-, intermediate- and short-term growth studies in asthmatic children treated
with inhaled glucocorticosteroids. European Respiratory lournal, 9, 821-827.
Wolthers, O.D., and Pedersen, S., 1990, Short term linear growth in asthmatic children during treatment
with prednisolone. British Medical Journal, 301, 145-148.
Wolthers, O.D., Hansen, M., luul, A., Nielsen, H.K., and Pedersen, S., 1997, Knemometry, urine
cortisol excretion, and measures of the insulin-like growth factor axis and collagen turnover in
children treated with inhaled glucocorticosteroids. Pediatric Research 41, 44-50.
CHAPTER 26
W. FURMAGA-JABLONSKAt, H. CHRZASTEK-SPRUCHt, M.
KOZLOWSKAt, A. ORZECHOWSKI:j:
1. INTRODUCTION
333
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 333-340.
© 2001 Kluwer Academic Publishers.
334 CHAPTER 26
3. RESULTS
The examined three groups of LBW children differed as for their gestational age. The
mean gestational age was 33.5 weeks for AGA preterms (group I), 35.8 weeks for S-
f-D newborns (group II), and 35.0 weeks for SGA preterms (group III). The
differences found for gestational age between S-f-D infants and AGA preterms as well
as for S-f-D infants and SGA preterms were highly statistically significant (ANOVA
and Duncan test, p < 0.001).
The biggest head circumference at birth and in the first two months of life was
revealed by S-f-D infants. After birth, mean differences between AGA preterms
(group I) and S-f-D infants (group II) amounted 1.5 cm, and between S-f-D infants
(group II) and SGA preterms (group III) - 1 cm. The differences were highly
significant for group I and II (p < 0.001) and for group II and III (p < 0.05),
respectively. Statistical significance was still observed in the first and second month
after birth, however, in later months of observation the differences for head
circumference between the groups were decreasing and were not statistically
significant (Figure 1).
In the subsequent months of the follow-up, a change of the rate of head
circumference velocity in individual groups of LBW children occurred. Between 3
months and 2 years (chronological age), the largest velocity of head circumference
was observed in AGA preterms (group I) while S-f-D infants (group II) tended to
have the lowest velocity. Between the 6th and 9th month, statistically significant
HEAD GROW11I AND PSYCHOMOTOR DEVELOPMENT 335
differences occurred for mean values of head circumference velocity between groups I
and II as well as between I and III (p < 0.05) (Table 1).
Development Quotient (DQ) in the three groups of LBW children differed from
one another in subsequent months of the investigation (Table 2). On the first
psychological examination, i.e. at 3 months, the lowest developmental level was
presented by AGA preterms, obtaining DQ = 84, which is below the 'normal' range.
Slightly higher mean score, DQ = 89, but still below the cut-off value of 90, was
found in the SGA preterms. Only S-f-D infants had, at 3 months, a DQ = 96, which
is within the 'normal' range. AGA preterms reached the developmental standard level
of psychomotor development of DQ = 90 by the age of 6 months month, while
mean DQ values in S-f-D newborns (DQ = 98) and in SGA preterms (DQ = 96)
exceeded this value. Differences between general developmental quotients in the 3rd
and 6th month of the longitudinal study for group I - AGA preterms and II -S-f-D
infants were statistically significant (in the 3rd month: p < 0.001, in the 6th month
p < 0.05).
Table 2. Mean values of developmental quotient (DQ) in the three groups of LBW children
52
48
E
u 44
~.
~ .: .:
~ 40
L*~iLI ~ L~p,,"'m, ll
§
~
'u
1"l
r "', ~ ~
36
<l)
::r: ~ --0-- ~
S-f-D infants
32 ~ ~ ...... j.........
mmmmrmmmmrmm : -------(r- SGA preterms .. ji.
28 I ;
-3 o 3 6 9 12 15 18 21 24
Months offollow-up
Figure 1. Mean head circumference the three groups of LBW children (* significant
diffe rences)
A similar tendency of changes in the three LBW groups was also observed in
anthropometric examinations. AGA preterms, despite the smallest mean head
circumferences at birth, were characterised by the highest growth rate of this trait
during the 24 month observation, while the smallest rate was found in S-f-D infants
338 CHAPTER 26
(Figure 1, Table 1). The above observations suggest correlations between physical
growth and psychomotor development which made the authors calculate Pearson's
correlation coefficient between head circumferences in individual months, head
circumference velocity expressed in cm/month and mean developmental quotients
obtained in the consecutive months of follow- up.
The correlations between head circumferences and psychomotor development for
AGA preterms are shown in Table 3. The highest correlations were found between
head circumference at birth and the general developmental quotient of AGA preterms
in the 6th (r = 0.57, p < 0.05) and in the 18th (r = 0.58, p < 0.05) month of their
lives. In AGA preterms, general developmental quotients in the 24th month of life
most strongly correlated with head circumference in the 2nd month of the
chronological age (r = 0.5, p < 0.05). Slightly lower correlation coefficients (r =
0.28 - 0.44), but still statistically significant (p < 0.05) were observed in AGA
preterms between the majority of physical and psychomotor parameters.
In the S-f-D infants, no significant correlations were observed between
psychomotor development in the 6th and 12th month of life and parameters of
physical growth (Table 3). However, general developmental quotient of S-f-D
children in the 3rd, 18th and 24th month of the chronological age were directly
proportional to head circumference in the first month of life (r = 0.52 - 0.59, p <
0.05). Correlations in S-f-D children occur for head circumference measurements
obtained in the 2nd month and DQ in the 9th, 18th and 24th month (r = 0.44 - 0.48,
p < 0.05). Psychomotor development in the 9th month also strongly correlates with
head circumferences of S-f-D children in the 2nd, 3rd and 6th month (r = 0.44 - 0.50,
p < 0.05).
The least significant correlations between head circumferences and psychomotor
development were found in group III of SGA preterms (Table 3). Statistically
significant correlations occurred in SGA preterms only between head circumference
in newborns and in the 1st month, and DQ in the 3rd and 6th month respectively (r =
0.48, r = 0.79, p < 0.05).
The velocity rate of head circumference did not correlate with DQ in the AGA
preterm group as well as the S-f-D infants. However in SGA preterms, we observed
highly significant correlation between velocity of the head circumference in the first
month and DQ in the 6th month of chronological age (r = 0.94, p < 0.05).
4. DISCUSSION
Several investigations of the literature have confirmed a close relationships between
the increase of head circumference and the brain development in normally developed
children during the first two years of their life (Simon et at. 1993, Brandt 1978,
Mauser 1984, Dobbing 1974). Based on the literature, we can assume that the
biggest development of brain starts in the mid-pregnancy and reaches the extreme
values in the 2-3rd year of life (Brandt 1978, Mauser 1984, Dobbing 1970, Dobbing
and Sands 1978, Winnick and Rosso 1969). This process is parallel to the biggest
increase of head circumference. In those periods of life the increase of brain mass and
the head circumference are connected with the development and fast progress of
psychomotoric functions in the youngest children. According to Henneberg et at.
(1984) the intelligence quotient is proportional to the efficiency at which the brain
processes and stores the information, and the bigger number of nervous elements is
HEAD GROWTH AND PSYCHOMOTOR DEVELOPMENT 339
placed in it, the bigger efficiency it has. The above considerations concern AGA full-
term children who present a regular biological development at the birth and whose
further physical growth and psychomotoric development is realised according to the
definite standards (Kurniewicz-Witczakowa 1980). In the available references no
attempts were made to define the relationships between the physical and
psychomotoric development in the group of children born with low-birth-weight.
However the investigations carried out in the world confirmed that the physical
growth and psychomotoric development between the LBW children and the AGA full
term-children differs substantially (Chrzastek-Spruch et al. 1998, Gajewska et al.
1995, Kaliszewska-Drozdowska 1989).
Earlier observations indicated that the existing differences in the dynamics of head
growth in the three groups of children, born with a low-birth-weight, influence the
different advancement of this feature in the second year of life (Furmaga-Jablonska et
al. 1997, Furmaga-Jablonska et al. 1999). Therefore, we made an attempt to define
possible relationships between the head circumference and psychomotoric
development in AGA preterms, S-f-D infants and SGA preterms, in view of
predicting the psychomotoric development in children on the basis of longitudinal
observations of head circumference development.
When we consider the relationships between physical and psychomotoric
development, the biggest relationships were confirmed between those two processes
in the AGA preterms. Statistically significant correlations were confirmed between
the head circumference at birth and the psychomotoric development in the subsequent
months of investigations starting from the 3rd till the 18th month of life. The
psychomotoric development (defined by general developmental quotient DQ) of AGA
preterms in the second year of life is also directly proportional to their head
circumferences in the 2nd, 3rd, 6th, 9th and 12th month of observation; but the
correlation coefficient decreases with the age of the child. Fewer correlations were
confirmed between the head circumference and psychomotoric development in the
group of S-f-D infants in the subsequent months of life. However the measurements
of head circumference performed in the 1st and 2nd month of chronological age are
diagnostically significant for the level of psychomotoric development in this group
in the second year of life.
In SGA preterms only measurements of head circumference at birth and in the
first month of life are statistically significant for prediction of their psychomotor
development in the 3rd and 6th month. And thus the newborns with double
pathology, i.e. pre-maturity and intrauterine growth retardation (IUGR) are
characterised by dissonance of further physical growth and psychical development.
However only in SGA-preterms a very direct correlation was confirmed between the
head circumference in the first month of life and the developmental quotient levels
(DQ) in the 6th month of life.
Considering the relationships between head circumference and psychomotoric
development in early childhood which are indicated in this paper, it seems that the
measurements of head circumference during developmental period may be a potential
screening test for evaluating the psychomotoric disorders in the group of low-birth-
weight children.
340 CHAPTER 26
5. REFERENCES
Brandt, I., 1978, Growth dynamics of low-birth-weight infants with emphasis on the perinatal period. In
Human growth, edited by F. Falkner and J.M. Tanner (New York: Plenum Press 2) p. 415.
Brandt, I., 1986, Patterns of early neurological development. In Human growth, edited by F. Falkner
and J.M. Tanner (New York: Plenum Press 2) p. 469.
Brunet, 0., Lezine, 1.,1951, Le developpement psychologique de la premiere enfance. Paris.
Chrzastek-Spruch, H., Jablonska, R, Furmaga-Jablonska, W., Kozlowska, M., Billewicz-Kraczkowska,
A., Hauspie, R, Orzechowski, A., 1998, Rozw6j psychomotoryczny wczesniak6w - badania
dlugofalowe. In Skad idziemy, kim jestesmy, dokad zmierzamy, edited by D. KornafeI (Wroc1aw:
Polskie Towarzystwo Antropologiczne) p. 33.
Dobbing, J., 1970, Undernutrition and the developing brain. American Journal of Diseases of Children,
120,411-415.
Dobbing, J., 1974, The later development of the brain and its vulnerability, In Scientific Foundation of
Paediatrics, edited by J.A. Davies and J. Dobbing, (London: Heinemann Medical Books) p. 565.
Dobbing, J., Sands, J., 1978, Head circumference, biparietal diameter and brain growth in fetal and
postnatal life. Early Human Development, 2, 81-87.
Furmaga-Jablonska, W., Chrzastek-Spruch, H., Kozlowska, M., 1999, Longitudinal analysis of head
growth in children born with low birth weight. Perspectives in Human Biology, edited by M.
Henneberg, Centre for Human Biology 4, 23-32.
Furmaga-Jablonska, W., Chrzastek-Spruch, H., Kozlowska, M., Jablonska, R, Orzechowski, A., 1998,
Rozw6j psychomotoryczny wczesniak6w w drugim roku zycia. In Czlowiek wczoraj, dzis, jutro,
edited by A. Gasiorowski, 1. Gurba and W. Kozak-Zychman, (Lublin: Wydawnictwo UMCS)
p.181.
Furmaga-lablonska, W., Kozlowska, M., Kulik-Rechberger, B., 1997, Dynamika rozwoju obwodu
glowy w trzech grupach dzieci urodzonych z mala masa ciala. Pediatria Polska, LXXII, 12, 1127-
1136.
Gajewska, E., Fuchs, B., Paluszynska, D., 1995, Noworodek 0 malej masie ciala, problem nadal
aktualny. (Warszawa: Centrum Zdrowia Dziecka)
Henneberg, M., Budnik, A., Pezacka, M., Puch, A., 1984, Wielkosc glowy a inteligencja:
wsp61zaleznosci wewnatrzgatunkowe u czlowieka, Przeglad Antropologiczny, 50, 299-313.
Kaliszewska-Drozdowska, M.D., 1989, Odrebnosci rozwojowe roznych grup niemowlat - propozycje
metodyczne oceny ich rozowju fizycznego., Pediatria Polska, 64,479-481.
Kurniewicz-Witczakowa, R, 1980, Rozw6j somatyczny, In Diagnostyka rozwoju dzieci i mlodziezy,
edited by 1., Kopczynska-Sikorska (Warszawa: PZWL) p.l20.
Lindhal, E., 1987, Motor performance of neonatal risk and non risk children at early school age, Acta
Paediatrica Scandinavica, 76, 809-817.
Livshits, G., Cohen, Z., Ortemski. I., 1992, Interrelations between early child development, gestational
age and birth weight, International Journal of Anthropology 2, 41-51.
Livshits, G., Cohen, Z., Ortemski, I., 1993, Relationship between physical growth and motor
development in infancy and early childhood: Multivariate analysis. American 10urnal of Human
Biology, 5, 481-489.
Mauser, 1.1., 1984, Growth in the high - risk infant, Clinics in Perinatology, 2,19-40.
Moczko, J.A., Breborowicz, G.H., and Tadeusiewicz, R, 1998, Statystyka w badaniach medycznych
edited by Springer PWN, Warszawa.
Ounstedt, M., Moar, V.A., Scott, A., 1984, Association between size and development at four years
among children who were small-for dates and large-for dates at birth, Early Human Development,
9,259-268.
Piekkala, P., Kero, P., Sillanpaa, M., 1988, The developmental profile and outcome of 325 unselected
preterm infants up to two years of age., Neuropediatrics. 19, 33-40.
Sheth, RD., Mullett, M.D., Bodensteiner, J.B., 1995, Longitudinal head growth in developmentally
normal preterm infants. Archives of Pediatrics and Adolescent Medicine 149, 1358-1361.
Simon, N.P., Brady, N.R, Stafford, RL., 1993, Catch-up head growth and motor performance in very-
low-birthweight infants. Clinical Pediatrics, 7, 405-411.
Winnick, M., Rosso, P., 1969, The effect of severe early malnutrition on cellular growth of human
brain, Pediatric Research, 3, 181-184.
CHAPTER 27
1. INTRODUCTION
In India thalassemia trait varies between 1 and 17 percent with an average of 3.3
percent in normal population and Hb-E disease is exclusively present in the
population of West Bengal and Assam (Chatterjee 1965, Talukder and Sharma
1994). It co-exists with beta-thalassemia in West Bengal producing serious public
health problems. Growth retardation occurs in homozygous beta-thalassemia at
about the age of 8-1 0 years and many affected children attain a very short final height
(Weatherall and Clegg 1981). Various auxological studies indicate that stunted
growth and delayed puberty is frequent in transfusion-dependent thalassemics
(Borgna-Pignatti et al. 1985, Perignon et al. 1993, Yeslipek et al. 1993; Saka et al.
1995). In India, where most of the thalassemic children are under-transfused and do
not get adequate chelation therapy (Verma et al. 1992), the age of onset of growth
retardation may not be the same as that of the developed countries. In the present
communication an attempt has been made to report growth and growth increments
in length/height and weight of thalassemic Bengali children aged 1-6 years in
relation to the non-thalassemic control.
341
P. Dasgupta and R. Hauspie (eds.), Perspectives in Human Growth, Development and Maturation, 341-349.
© 2001 Kluwer Academic Publishers.
342 CHAPTER 27
Forty-five percent of the patients participating in the study came from various
districts of West Bengal whereas 42% were residents of the city of Calcutta. The
patients were predominantly low caste Hindus (75%). Twenty-three percent parents
of the affected families were recorded as illiterate while 15% possess higher
education. Fifty-seven percent of the patients were found to represent the families
from service holders. However with respect to the per capita annual income level
(Rs. 9,000) economic status of the patients appeared to be low with average family
size of 5.2. Seventy-one percent of the children below six years of age showed
evidences of malnutrition in varying degrees whereas the rest (29%) appeared to be
normal. Mortality rate among the thalassemics were 12.12%.
Normal children were examined by the expert paediatricians and they were non-
thalassemics without having any genetic disorder. Among them about 39% stayed
in the city of Calcutta while the rest were from various districts of the state of West
Bengal. Eighty-four percent of the them comprised the higher castes like Brahmin,
Vaidya and Kayastha while the rest were affiliated to low or scheduled castes. Five
percent of the mothers were illiterate while 58% had education up to the school level
and 37% had education up to the college level. With respect to the occupation of
father only 16% were engaged in Agriculture, 49% were service holders, 35% were
involved in small scale business. The average family size was 3.5 with the per
capita annual income ofRs. 12,510.
n Nbr. of occasions
2 3 4 5 6
thalassemic height (boys) 81 34 17 10 10 1 9
thalassemic height (girls) 45 9 15 8 5 2 6
thalassemic weight (boys) 81 34 17 10 10 1 9
thalassemic weight (girls) 45 9 15 8 5 2 6
non-thalassemic height (boys) 41 3 5 12 4 3 14
non-thalassemic height (girls) 30 1 2 2 4 5 16
non-thalassemic weight (boys) 41 3 5 12 4 3 14
non-thalassemic weight (girls) 30 1 2 2 4 5 16
A mixed-longitudinal method of growth study was adopted. From the first visit
onward each child of non-thalassemics and thalassemics category was followed up
till six years. However, there were several drop outs which had been replaced by the
new entrants into the study. The frequency of measurements made on the
thalassemic children and the controls, by sex and variables are shown in Table 1.
The pure longitudinal element of the data will be treated in a subsequent analysis.
The children visited the blood transfusion units as and whenever their
haemoglobin fell to a level when they physically felt very ill. During such visits
they were measured for lengthlheight and weight but only the measurements taken
on their birth dates were used in the present communication. Thus the ages refer to
exact ages.
GROWTH OF THALASSEMIC CHILDREN 343
All children were measured for lengthiheight and weight at forenoon following
the recommendation of Gam and Shamir (1958) and the measurements were taken
following Martin and Saller (1957). Supine length of the infants was taken up to the
nearest centimetre till 2 years of age with the help of an infantometer specially
designed for this purpose. Measurements were taken by one of us (JB). Body weight
was measured with a portable weighing machine and height was measured with an
anthropometric rod.
For clinical diagnosis of the thalassemics (i) estimation of haemoglobin, PCV,
MCV, reticulocyte count, (ii) agarose gel electrophoresis at pH 8.6 for detection of
haemoglobin variants, (iii) foetal haemoglobin by alkali denaturation and (iv) HbA2
by gel elution technique (Dacie and Lewis 1990) were performed. During the follow-
up period, pre-transfusion haemoglobin level was estimated for each child.
Data for length/height and weight of the thalassemics and the controls were
analysed by the method of Tanner (1951) applied for the mixed-longitudinal design.
Age specific means and variances for lengthiheight and weight of the thalassemics
were compared with the controls. Mean and variances of the increments have been
calculated on the same data by Tanner's (1951) method. Statistical significance of
the mean difference between the thalassemics and the non-thalassemics in all ages in
attained size as well as increments were tested by the unpaired Student's t-test. A p-
value of less than .01 have been considered to be significant.
3. RESULTS
Age and sex specific means and variances of lengths/heights and weights of the
thalassemic and non-thalassemic children are presented by age and sex in Tables 2
and 3. The means refer to the M h' calculated according to the formula of Patterson
(1950), an estimator of central tendency extracting most of the information contained
in mixed-longitudinal growth data (Tanner 1951).
In both traits and for both sexes the thalassemics showed lower mean values than
the non-thalassemics controls. For height, in both sexes, the mean difference
increased from one year and continues till 5 years after which it declined. For weight,
the increasing trend continued in all ages for boys excepted for the girls only at 6
years. As a whole, both thalassemic and non-thalassemic boys showed larger means
than the girls.
The variances in height for both sexes of the thalassemics were relatively larger
during the earlier ages which declines in mid ages and finally increased in the
subsequent years. On the contrary, variation in weights over different ages for both
sexes was more or less consistent.
The non-thalassemics, however, did not exhibit any consistent trend in variances
with respect to height, but in weight they exhibited the same consistency as in the
thalassemics. Statistically significant differences were observed between the means of
the thalassemics and non-thalassemics at all ages in both traits (p<.0 1).
One-year increments were calculated as follows: the increment between 1 and 2
years is assigned to the age 1.5 years, i.e. the centre of the interval (see Tables 4 and
5). The mean increments of lengthiheight of the thalassemics, as expected, were less
than the non-thalassemics, except at 5.5 years among the girls where a reversal was
noticed. The maximum difference in the mean increments between the thalassemics
344 CHAPTER 27
and the non-thalassemics was at age 1.5 in boys. However, among the girls the
maximum difference was at 4.5 years. Within the thalassemics group, in both sexes,
the maximum decline of the increments in height was observed after 3.5 years. For
weight, the difference in mean increments between the thalassemics and non-
thalassemics was maximum at 1.5 years (in both sexes). Further, in weight within
the thalassemics group the maximum decline in increment was noticed after 2.5
years of age.
Table 2. Means and variances of length/height (em) of the thalassemics and non-
thalassemics (normal controls) by age and sex. All differences were statistically
significant (t-Student p<.OJ)
thalassemics non-thalassemics
Age, years n mean variance n mean variance
Boys
1 23 71.59 1.85 20 73.50 0.36
2 22 77.98 1.79 21 82.83 0.64
3 37 83.56 0.96 29 90.25 1.11
4 37 90.14 0.76 30 97.54 1.06
5 40 95.66 0.76 33 105.08 0.54
6 38 101.47 0.81 31 110.45 0.25
Girls
1 13 68.50 3.71 16 73.00 0.68
2 21 76.27 1. 78 21 82.00 0.25
3 25 81.16 0.97 25 88.50 0.37
4 22 86.82 0.96 27 95.52 0.92
5 27 92.31 1.24 29 103.50 0.77
6 21 97.84 2.22 30 108.52 0.74
Table 3. Means and variances of weight (kg) of the thalassemics and non-thalassemics
(normal controls) by age and sex. All differences were statistically significant (t-
Student p<.OJ)
thalassemics non-thalassemics
Age, years n mean variance n mean variance
Boys
1 24 7.44 0.04 20 8.50 0.03
2 23 8.86 0.04 21 10.97 0.04
3 37 10.84 0.04 29 13.58 0.08
4 37 11.84 0.06 30 14.91 0.06
5 40 13.11 0.06 33 16.64 0.07
6 38 14.46 0.11 31 18.81 0.07
Girls
1 13 7.03 0.09 16 8.01 0.06
2 21 8.51 0.09 21 10.52 0.04
3 25 10.56 0.05 25 12.75 0.04
4 22 11.34 0.07 27 14.00 0.04
5 27 12.86 0.10 29 15.75 0.06
6 21 14.21 0.19 30 17.00 0.07
GROWTH OF THALASSEMIC CHllDREN 345
thalassemics non-thalassemics
Age, years n mean variance n mean variance
Boys
1.5 22 5.65 0.53 21 9.09 0.32
2.5 37 5.05 0.45 29 7.54 0.35
3.5 37 5.71 0.54 30 7.93 1.30
4.5 40 4.80 0.42 33 7.79 0.42
5.5 38 4.79 0.38 31 6.11 0.18
Girls
1.5 21 6.41 1.01 21 8.72 0.16
2.5 25 4.45 0.40 25 6.15 0.08
3.5 22 5.32 0.79 27 6.70 0.25
4.5 27 3.97 0.30 29 9.31 0.29
5.5 21 4.49 0.18 30 3.98 0.14
thalassemics non-thalassemics
Age, years n mean variance n mean variance
Boys
1.5 23 1.36 0.02 21 2.67 0.02
2.5 37 1.47 0.04 29 2.61 0.04
3.5 37 0.70 0.02 30 1.58 0.02
4.5 40 0.97 0.03 33 1.74 0.03
5.5 38 1.09 0.03 31 2.27 0.02
Girls
1.5 21 1.36 0.05 21 2.51 0.01
2.5 25 1.80 0.06 25 2.18 0.01
3.5 22 0.66 0.06 27 1.34 0.004
4.5 27 1.10 0.03 29 1.69 0.01
5.5 21 0.94 0.03 30 1.24 0.01
4. DISCUSSION
In the present study the thalassemic group of children have displayed early onset of
reduced size and delayed growth (from age 1 year) and behaved like untreated
patients most probably due to inadequacy and irregularities in taking blood
transfusion. It has been suggested by several workers that retardation of various
parameters at an early age are more likely to be secondary to chronic hypoxia,
followed by inadequate blood transfusion therapy (George et al. 1997). Some
workers (Fuchs et al. 1996, Benso et al. 1990) have remarked that while hyper-
transfusion regimen can modify the pattern of growth disorders, retardation in growth
and development persists even in a transfused children.
120,
Boys
§ 110
~ 100
'iii
<E 90
~
'&
c:
~ 80 . . - Thalassemia
>- Non-Thalassemia
•• _. & •••
'0 ~ICMR
0 70 ~NCHS
CD
60 0
25 ~
Boys
I
20
OJ
~
1: 15
Ol
'iii
==
>- 10 . . - Thalassemia
'0
0 Non-Thalassemia
- - - -&. - •
CD
5 &--ICMR
)( NCHS
00
Figure 1, Mean length (top) and weight (bottom) of thalassemic and non-thalassemic
boys, compared with the 50th centile of the Indian (ICMR 1972) and the NCHS
reference (Hamill et at. 1979)
Our results are in variance with the study by Johnston et at. (1966), and
Piomelli et ai, (1974) who found that growth retardation of the thalassemic children
from four years onwards only. In a study by Kattamis and Kattamis (1995), it was
reported that untreated patients with thalassemia major (type 1 and 2) showed
GROWfH OF THALASSEMIC CHILDREN 347
growth retardation during the first decade of life and were below the 3rd centile for
weight and height in comparison to the non-thalassemic children. In a recent Indian
study, carried out in New Delhi (George et al. 1997), the authors did not fmd any
significant difference between the tha1assemics and the non-thalassemics children in
average height. till 9 years. The study, however, has not provided any estimate of
growth increments.
120 I
Girls
E 110
0
Cl 100
1:
'iii
.c
:cc;, 90
c
..!!? 80 ....-- Thalassemia
>. - - - -€>- - - - Non-Thalassemia
'0
~ICMR
co 70
0
)( NCHS
60 0
2 3 4 5 6 7
Age, years
25, Girls
20
Cl
.:.!
1: 15
Cl
'iii
~
>.
10
'0 ....-- Thalassemia
0
co - - - - €>- - - - Non-Thalassemia
5 ~ICMR
)( NCHS
00 7
Figure 2, Mean length (top) and weight (bottom) of thalassemic and non-thalassemic
girls, compared with the 50th centile of the Indian (ICMR 1972) and the NCHS
reference (Hamill et at. 1979)
Age-specific mean heights and weights of the thalassemic children of both sexes
show lower scores than the 50th centiles of the ICMR (Indian Council of Medical
Research 1972) and the NCHS (Hamill et al. 1979) standard (Figures 1 and 2).
Moreover, it has been observed that mean body weight of the thalassemic children
corresponds to the 25th percentile of the ICMR standard at all ages. But for height
the corresponding ICMR percentile was less than 10. Thus it clearly indicates that
the children of the present study were shorter with respect to their weight. Overall,
the thalassemic children fell below the 5th percentile of the NCHS standard (Hamill
et al. 1979) and showed stunted growth from the age of one year.
348 CHAPTER 27
A number of studies have now established that stunted heights of the thalassemic
children are due to disproportionate growth between the upper and lower segments,
sitting height or trunk height being more affected than leg length (Rodda 1994,
Fuchs et al. 1997). Unfortunately, the present study does not provide any
opportunity to examine growth velocity of the two principal body segments.
The causative factors which have been found to be responsible for growth
retardation of the thalassemics, may be summarised as follows:
(i) Abnormal Hypothalamus Pituitary Gonadal axis function (Costin et al. 1979).
(ii) Decrease in the production oflGF -1 (Herington et al. 1981).
(iii) The toxic effects of Desferrioxamine (de Virgilis et al. 1988) (though not
applicable in the present study).
(iv) Cellular hypoxia, due to anaemia (Fuchs et al. 1996).
(v) Deficiencies in the micronutrients (Arcasoy and Cavdar 1975).
(vi) Undernutrition, due to inadequate nutrient intake (Fuchs et al. 1997).
(vii) Chronic liver disease (de Sanctis et al. 1991).
However the results from some of these investigations are inconsistent or
contradictory (Fuchs et al. 1997).
The present study however was not aiming to investigate the role of these factors
for the causation of the growth retardation in this heterogeneous thalassemic children
with small sample size. Heterogeneity might have been caused due to different
clinical histories of the patients. Constantoulakis et al. (1975) have commented that
low haemoglobin level and severity of the disease may hinder normal growth of the
thalassemic patients. Nevertheless, it is one of the first mixed-longitudinal study on
the thalassemic children of India which has generated some basic growth data.
Future studies with larger sample size should be carried out to disentangle the
relative role of these factors resulting in growth retardation of the Indian thalassemic
children.
5. REFERENCES
Arcasoy, A., and Cavdar, A.C., 1975, Changes of trace minerals (serum iron, zinc, copper and
magnesium) in thalassemia. Acta Haematologica, 53, 341-345.
Benso, L., Pastorin, L., Gambotto, S., Petri, A., Gianino, P., Signorile, F., and Gabutti, V., 1990,
Auxological Study of the onset of puberty and of the age at menarche in B-thalassemic patients.
Acta Medica Auxologica, 22, 51-56.
Borgna-Pignatti, C.D., Stefano, P., Zonta, L., Vullo, C., de Sanctis, V., Melevendi, C., Naselli, A.,
Masera, G., Tarzoli, S., and Gabutti, V., 1985, Growth and sexual maturation in thalassemia major.
American Journal of Physical Anthropology, 106, 150-155.
Chatterjee, J.B., 1965, Some aspects of haemoglobin E and its genetic interaction with thalassemia.
Indian Journal of Medical Research, 53, 377-398.
Constantoulakis, M., Panagopoulos, G., and Augoustaki, 0., 1975, Stature and longitudinal growth in
thalassemia major. Clinical Pediatrics, 14, 353-368.
Costin, G., Kogut, M.D., Hyman, C.B., and Ortega, J.A., 1979, Endocrine abnormalities in thalassemia
major. American Journal of Disease in Childhood, 133,497-502.
GROWTH OF THALASSEMIC CHILDREN 349
Dacie, J.V., and Lewis, S.M., 1990, Investigation of the abnormal haemoglobins. In Practical
Haematology (England: Churchill, Livingstone).
de Sanctis, V., Atti, G., Banin, P., Orzincolo, C., Cavallini, A.R, Patti, D., and Vullo, C., 1991, Growth in
thalassaemia major. Acta Medica Auxologica, 23, 29-36.
de Virgilis, S., Congia, M., Frau, F., Argiolu, F., Diana, G., Cucca, F., Varsi, A., Sanna, G., Podda, G.,
Fodde, M., Pirastu, G.F. and Cao, A., 1988, Desferrioximine-induced growth retardation in patients
with thalassaemia major. Journal of Paediatrics, 113,661-669.
Fuchs, GJ., Tienboon, P., Khaled, M.A., Nimsakul, S., Linpisarn, S., Faruque, A.S.G., Yutrabootr, Y.,
Dewier, M., and Sus kind, RM., 1997, Nutritional support and growth in thalassaemia major.
Archives of Disease in childhood, 76, 509-512.
Fuchs, GJ., Tienboon, P., Linpisarn, S., Nimsakul, S., Leelapat, P., Tovanabutra, S., Tubtong, V.,
Dewier, M., and Suskind, R.M., 1996, Nutritional factors and thalassemia major. Archives of
Disease in Childhood, 74, 224-227.
Gam, S.M., and Shamir, S., 1958, Methods of Research in Human Growth (Springfield, Illinois: e.e.
Thomas).
George, A., Bhaduri, A., Sen, S., and Choudhry, V.P., 1997, Physical growth parameters in thalassemic
children. Indian Journal of Pediatrics, 64, 861-871.
Hamill, V.V., Drizd, T.A., Johnson, C.L., Reed, RB., Roche, A.F., and Moore, W.M., 1979, Physical
Growth: National Centre for Health Statistics Percentiles. The American Journal Of Clinical
Nutrition, 32, 607-629.
Herington, A.C., Werther, G.A., Matthews, RN., and Burger, H.G., 1981, Studies on the possible
mechanism for deficiency of non-suppressible insulin-like activity in thalassemia major. Journal of
Clinical Endocrinology and Metabolism, 52, 393-398.
Indian Council of Medical Research, 1972, Growth and Physical development of Indian infants and
children. Technical report series no. 18 (New Delhi: ICMR).
Johnston, F.E., Hertzog, K.P., and Malina, RM., 1966, Longitudinal growth in thalassemia major.
American Journal of Disease in childhood, 122,396-403.
Kattamis, e.A., and Kattamis, A.C., 1995, Management of thalassemias: Growth and development,
honnone substitution, vitamin supplementation and vaccination. Seminars in Haematology, 32, 269-
279.
Martin, R, and Saller, K., 1957, Lehrbuch der Anthropologie. (Germany: Jena, Gustav- Fischer).
Patterson, H.D., 1950, Sampling on successive occasions with partial replacement of units. Journal of
the Royal Statistical Society, 12,241-255.
Perignon, F., Braunar, R, Souberbielle, J.e., de-Montalembert, M., and Girot, R., 1993, Growth and
endocrine function in major thalassemia. Archives of French Paediatrics, 50, 657-663.
Piomelli, S., Karpatkin, M.H., and Arzanian, M., 1974, Hypertransfusion regimen in patients with
Cooley's anemia. Annals of the New York Academy of Sciences, 232,186-192.
Rodda, C.P., 1994, Body proportions in patients with homozygous beta - thalassemia. Acta Pediatrica
(Supplement), 406, 107-108.
Saka, N., Sukur, M., Bundak, R., Anak, S., Neyzi, 0., and Sedikoglu, G., 1995, Growth and puberty in
thalassemia major. Journal of Pediatric Endocrinology and Metabolism, 8, 181-186.
Talukder, G., and Sharma, A., 1994, Haemoglobinopathies in India. Journal of Human Ecology, 3, 125-
130.
Tanner, lM., 1951, Some notes on the reporting of growth data. Human Biology, 23, 93-159.
Venna, I.C., Choudhry, V.P., and Jain, P.K., 1992, Prevention of thalassemia: A necessity in India.
Indian Journal of Paediatrics, 59, 649-654.
Weatherall, DJ., and Clegg, lB., 1981, The Thalassemia Syndromes (Boston: Blackwell Scientific
Publications) 3rd edition.
Yeslipek, M.A., Bircan, I., Oygur, N., Ertug, H., Vegin, 0., and Guven, A.G., 1993, Growth and sexual
maturation in children with thalassemia major. Haematologica, 78, 30-33.
ApPENDICES
Life sketch of Sudhir Ranjan Das
Sudhir Ranjan Das received his initial M.Sc. degree in Pure Physics from Calcutta
University in 1933. He worked as a Research Fellow in Biophysics at the Bose
Institute from 1942 to 1948. He developed special interest in radiology and was
appointed Radiologist-Anthropologist in the Anthropological Survey of India in
1948. It was a time when the Survey was in search of young talented individuals to
build a strong research department.
With this background and training, he specialised in human genetics and human
physical growth. He developed the radiology section of the Survey, and commenced
a longitudinal growth study among the Brahmin, the Kayastha and Vaidya
communities in the suburbs of Calcutta (Sarsuna-Barisha Growth Study). A
radiological study of the skeletal maturation formed the part of the project. Besides
being involved in this long-term project, Das explored a number of areas like
serology, somatology, PTC taste factor, sickle-cell traits and colour blindness,
distribution of LDH variants and placental alkaline phosphatase, enzyme, serum
proteins and Bart's haemoglobin survey. His works were not limited to Bengal or
Calcutta suburbs, rather he worked on populations in Orissa, Nagpur and Kerala as
well.
He retired from the Survey in 1968. After his retirement from the Survey he
joined the Indian Statistical Institute, Calcutta, as Professor and Head of the
Anthropometry and Human Genetics Unit (now Anthropology and Human Genetics
Unit) and continued his research activities until he left the Institute. Das knew four
languages, namely, Bengali, Hindi, Sanskrit and English.
Das had published extensively and at times it is difficult to prefer one above the
other. Nonetheless some of them are of special significance, namely, Inheritance of
PTC taste character in man: an analysis of 126 Rarhi Brahmin families of West
Bengal in Annals of Human Genetics (1958), Blood groups (ABO, M-N, Rh), ABH
secretion, sickle-cell, PTC taste and Colour Blindness in the Mahar of Nagpur, in
the Journal, Royal Anthropological Institute (1961), A spectrophotometric skin
colour survey among four Indian caste and Tribes, in Zeitschrift flir Morphologie
und Anthropologie (1963), Caste variations of lactate dehydrogenase in the
Dhangars of Maharashtra, India, in the Japanese Journal of Human Genetics
(1973), Blood groups, Serum proteins, haemoglobin, and some serum and red-cell
enzymes among the Kaoras of 24-Parganas in West Bengal, in Human Heredity,
(1974), and A Radiological Study of Skeletal Maturity in Hand-Wrist-Area of
Bengali Boys and Girls, in Anthropological Survey ofIndia, Calcutta (1976).
The major contribution of late Prof. Das are also given herewith chronologically:
conversion to the rhombic, and other properties were found out. The results
appeared mainly in Indian Journal of Physics, Calcutta and two papers in
Nature, London, 1937/39. J.R. Partington, London, Newton & Friend, from
Cambridge University, quoted the findings elaborately in the Advanced
Chemistry books. A paper was read and published in the Proceedings of the
International Union of Chemistry on sulphur and selenium in Munster
(Germany) 1956.
The entire work was done with new types of suitable X-ray diffraction cameras,
designed by him.
During 1940-47 he worked in the Bose Institute (Acharya J.C. Bose), Calcutta,
with a break of 18 months in Kabul University as a lecturer of Physics. In the
Bose Institute he took up:
(i) the construction of a powerful source of ultra-sonic generator for the Defence;
(ii) the study of protein structure by the X-ray diffraction method;
(iii) the induction of mutations in some plants by X-ray irradiation of seeds;
(iv) the construction of a powerful artificial source of neutron beam suitable for
use in medicine.
3 He worked in the Anthropological Survey, Government of India, from 24
October 1948 to 31 March 1968. The following lines drew his attention:
(a) X-ray study of hand-wrist skeletal (by epiphyseal union) and physical growth in
about 500 Bengali Boys and Girls - from birth to 21 years - in mixed
longitudinal samples (Sarsuna-Barisha Growth Study). Twenty-two
anthropometric characters of each child, measured on birth-days or half birth-
days, supplied the basic growth data.
The growth and maturity field study covered a period of 14 years - 4 April
1952 to 31 March 1966 - and all measurements were take by a single observer
(S.R.D.). The skeletal maturity was assessed on the Greulich and Pyle
Standards U.S.A. 1958. A report published by Anthropological Survey of
India, Government ofindia, Calcutta, 1970.
Professor J.M. Tanner (London) kindly contributed to the analysis of the
voluminous growth metric data in 1969-70. The data have since been only
partially analysed, by Prof. J.M. Tanner, Prof. R.C. Hauspie (Brussels), Prof.
S.R. Das, Dr. M.A. Preece and Dr. M.J. Baines (London). Several papers have
already been published in Annals of Human Biology and elsewhere. More
papers are getting ready for press.
Scholars from all over the world may have access to these materials on
application to the Director of the Anthropological Survey of India, Government
ofindia, Calcutta, India
(b) A paper in three parts appeared in Metron, Roma (1953-55) on "A Mathematical
analysis of the phenomena of human twins, triplets, and higher plural births, I,
II, III". This paper answered definitely several long-standing questions about
these phenomena, removing certain wrong concepts, and throwing new light on
them.
(c) S.R. Das introduced a correct methodology, developing an instrument,
dactylometer, introducing a scientific positioning of the hand-wrist-forearm on a
board and measuring the actual positions of the finger tips in this standard
position.
355
He studied a few castes and tribal populations for distribution of the manual
Digital Formulas in them.
The relative lengths and positions of the hand bones were investigated in
relation to the digital formula of the hand by X-ray siagrams in some 200 adult
men and women. A paper appeared in Zeitschrift fUr Morphologie und
Anthropologie.
(d) Phenylthiocarbamide (PTC) taste inheritance:
The single-locus di-allelic inheritance with complete dominance of the Taster
(T) gene seemed not to fit in with carefully collected PTC taste sibship or
family data (Harris & Kalmus, Annals of Human Genetics, London 1956-58).
Similar data obtained from some 125 Bengali families also indicated this
discrepancy. The hypothesis was modified by supposing that the T-gene was
not completely dominant in the heterozygotes (Tt) and as a result some of the
Tt individuals were non-tasters like (tt).
Papers appeared in Annals of Human Genetics, London 1956 & 1958 and
another in Proceedings of the Indian Anthropological Society, 1960, which
demonstrated satisfactory agreement of the incomplete dominance hypothesis
with his family data.
(e) Gene distribution in various Indian castes and tribes - the following gene
system were considered:
(i) Blood groups and types (AI, A2, B, 0, MN, Ss, Rh-C-D-E, Kell, Duffy,
Diego);
(ii) Secretion of ABH in saliva;
(iii) Sickle-cell;
(iv) red-green colour-blindness;
(v) PTC taste genes.
S.R. Das worked in the Indian Statistical Institute, Calcutta from 1968 (April)
to June 1973, when he retired from anthropology officially. Here he worked in
collaboration with Prof. B.N. Mukherjee of the Institute and Professor R.L.
Kirk of the Australian National University. In the Institute, he carried out the
following research activities:
He studied distribution of the genetic variants of haptoglobin and transferrin,
abnormal haemoglobins, the various red cell enzyme variants - the method
adopted was that of starch-gel Electrophoresis. In course of the studies two new
genetic variant ofLDH namely, Cal-l and Cal-2 were discovered. Hb-Barts was
detected in a sample of about 800 cord blood samples (1 %). Enzymes in human
placenta were investigated in the Calcutta hospitals and the results appeared in
Japanese Journal of Human Genetics. A number of papers appeared in Human
Heredity (Basel), in Human Genetik (Germany), etc ...
Das, S.R., 1936, A study of the structures of the various allotropes of Sulphur by
the X-ray Powder Photograph method. Science and Culture, vol. 1, June, 784-
785.
Das, S.R., and Ray, K, 1936, Structure of the allotropic Modifications of Sulphur.
Science and Culture, vol. II, August, 108-109.
Das, S.R., and Ray, K, 1937, The allotrope of Sulphur (alpha) study by the X-ray
diffraction method. Science and Culture, vol. II, June, 650-652.
Das, S.R., 1937, X-ray analysis of the structure of jute fibres. Science and Culture,
vol. II, June, 652-653.
Das, S.R., and Ghose, K, 1938, The structure of Sulphur particles in colloidal
suspension in water. Science and Culture, vol. IV, August, 132-133.
Das, S.R., 1938, A study of sulphur allotropes by the X-ray diffraction method. Part
1. Indian Journal of Physics, 12, 163-181.
Das, S.R., and Ghosh, K., 1939, A study of Sulphur allotropes by the X-ray
diffraction method (Part II). White Sulphur, Black Sulphur and Colloidial
Sulphur Suspensions in Water. Indian Journal of Physics, 13,91-105.
Das.S.R., and Dasgupta, K, 1939a, Conversion of Vitreous and monoclinic(alpha)
Selenium to the Hexagonal modification. Nature, 3613, 165.
Das,S.R., and Dasgupta, K., 1939b, X-Ray Diffraction by Supercooled Liquid
Sulphur. Nature, 143, 332.
Ray, B.B., Das, S.R., and Bagchi, N.1939. The K-absorption edges of cobalt in
cobalt metal and its compounds. Science and Culture. vol. IV, 528-529.
Ray, B.B., Das, S.R., and Bagchi, N., 1940, Secondary K-absorption edges of
Cobalt salts in solid and liquid solutions. Indian Journal of Physics, 14,37-54.
Roy, B.B., Dasgupta, K., Bose, H., and Das, S.R., 1940, The Flouroscence of
organic compounds by X-rays. Science and Culture, 5, 496-497.
1941 - 1950
Dasgupta, K., and Das, S.R., 1941, X-ray study of selenium in the liquid and
colloidial state. Indian Journal of Physics, 15,401-409.
Dasgupta, K, Das, S.R., and Ray, B.B., 1941, A study of allotropes of Selenium
by the X-ray diffraction method. Indian Journal of Physics, 15,389-399.
1951 - 1960
Das, S.R., 1952, The Human Manual digital formula and a standard hand axis - a
new method and an instrument foe their correct determination. Bulletin of the
Department of Anthropology Calcutta, 1, 105-114.
Das, S.R., 1954a, An instrument of direct determination of manual digital formula -
The dactylometer. Bulletin of the Department of Anthropology, Calcutta, 3, 152-
154.
357
358
Das, S.R., 1954b, A study of manual digital formula among the Bengalee and three
south Indian tribes. Bulletin of the Department of Anthropology, Calcutta, 3, 73-
85.
Das, S.R., and Ghose, L., 1954, A genetic Survey among the Paniyans. A South
Indian aboriginal tribe. ABO, MN, blood groups, secretor factor and taste
ability. Bulletin of the Department of the Anthropology, Calcutta, 3, 65-70.
Das, S.R., 1955, A somatological study of the Paniyans of Wyanaad. Bulletin of
the Department of Anthropology, Calcutta, 4, 51-61.
Das, S.R., and Mukherjee, D.P., 1955, A Roentgenometric Study of hand bones of
the Bengalee. Bulletin of the Department of Anthropology, Calcutta, 4, 2, 11-30.
Das, S.R., 1956a, A contribution to the heredity of the PTC taste characters based
on a study of 845 sib pairs. Annals of Eugenics, 20, 334-343.
Das, S.R., 1956b, A mathematical analysis of the phenomena of Human twins in
higher plural births (Refmed method). Part III. Metron, 18,63.
Das, S.R., 1958, Inheritance of the PTC taste character in man, An analysis of 126
Rarhi Brahmins families of West Bengal. Annals of Human Genetics, 22, 202-
212.
Das, S.R., and Sastry, D.B., 1960, Simplified maximum likelihood estimation of
the A-B-O gene frequencies: Results for a few Indian people surveyed. Bulletin of
the Anthropological Survey ofIndia, 9, 2, 1-8.
1961 - 1970
Das, S.R., Kumar, N., Bhattacharjee, P.N. and Satry, D.B, 1961, Blood Groups
(ABO, MN & Rh ), ABH secretion, sickle cell, PTC taste and colour blindness
in the Mahar of Nagpur. Journal of the Royal Anthropological Institute, London,
91, 345-355.
Mukherjee, D.P., and Das, S.R., 1961, Effect of Geological Aging on the particle
size of the bone mineral. Current Science, 30, 138-139.
Das, S.R., and Mukherjee, D.P., 1962, The anatomical basis of the human digital
formulas. Zeitschrift flir Morphologie und Anthropologie, 52, 163-175.
Das, S.R., Bhattacharjee, P.N., Sastry, D.B., and Mukherjee, D.P., 1962, Blood
groups (ABO, MN, Rh) ABH secretion, sickle cell trait and colour blindness in
the Bado Gadaba and Pareng Paroja of Koraput District of Orissa. Bulletin of the
Department of Anthropology, 11,324, 145-157.
Das, S.R., and Bhattacharjee, P.N., 1963, Blood groups (AI, A2, B) ABH
secretion, sickle cell, PTC taste and colour blindness in the Rajbanshi of the
Midnapur District, West Bengal. Bulletin of the Department of Anthropology,
12, (1 and 2), 1-6.
Das, S.R., and Mukherjee, D.P., 1963, A spectrophotometric skin colour survey
among four Indian castes and tribes. Zeitschrift fUr Morphologie und
Anthropologie, 54, 190-200.
Das, S.R., Mukherjee, D.P., and Bhattacharjee, P.N., 1963, PTC taste threshold
distribution in the Bado Gadaba and the Bareng Paroja of Koraput District. In
Orrissa. Acta Genetica et Satistica Medica, Basel, 13,369-377.
Das, S.R., and Mukherjee, D.P., 1964, Phenylthiocarbamide taste sensitivity
survey among the Pareng Gadaba, the OHaro Gadaba and Konda Paroja of
Korapur District, Orissa. Acta Genetica et Statistica Medica, Basel, 14, 168-176.
359
Das, S.R., 1966, Genetics of PTC taste character in Man - incomplete dominance
and the penetrance of the taste gene. In: Human Adaptability to environments
and physical fitness. M.S. Malhotra (ed). Defence Institute of Physiological and
Allied Sciences, Madras, Research and Developmental Organisation. Ministry of
Defence, Government ofIndia, 250-258.
Das, S.R., Sastry, D.B., and Mukherjee, D.P., 1966, Blood Groups (AI, A2, BO,
MN, Rh) and ABH secretion in the Pareng Ollaro Gadaba and the Konda Paroja
of Koraput District in Orissa. Acta Genetica et Statistica Medica, Basel, 16, 179-
183.
Das, S.R., 1966, Application of phenylthiocarbamide taste character in the study of
racial variation. (data on World taste gene distribution). Journal of the Indian
Anthropological Society, Calcutta, 1,63-80.
Das, S.R., Mukherjee, D.P., and Sastry, D.B., 1967, Sickle cell trait in Koraput
district and other parts of India. Acta Genetica et Statistica Medica, 17, 62-73.
Das, S.R., Mukherjee, D.P., and Bhattacharjee., P.N., 1967, Survey of the Blood
Groups and PTC taste among the Rajbanshi caste of West Bengal. (ABO, MNS,
Rh, Duffy and Diego). Acta Genetica, Basel, 17,433-445.
Das, S.R., Mukherjee, D.P., and Sastry, D.B., 1968, A somatological Survey of
the five tribes in the Koraput district, Orissa. Bulletin of the Department of
Anthropology, Calcutta, 17,4, 400-422.
Das, S.R., Mukherjee, B.N., and Das, S.K., 1970, The distribution of some
enzyme systems among Bengalis. The Indian Journal of Medical Research, 58,
866-875.
Das, S.R., Mukherjee, B.N., Das, S.K., Blake, N.M., and Kirk, R.L., 1970,
Placental alkaline Phosphatase types in Calcutta, India. Japanese Journal of
Human Genetics, 3,155-158.
Das, S.R., Mukherjee, B.N., Das, S.K., Ananthakrishnan, R., Blake, N.M., and
Kirk, R.L., 1970, LDH variants in India. Humangenetik, 9, 107-109.
1971 - 1980
Das, S.R., Mukherjee, B.N., and Das, S.K., 1971, Genetic variants of Lactate
Dehydrogenese in India. Journal of the Indian Anthropological Society, 6, 2,
135-144.
Das, S.R., Mukherjee, B.N., and Das, S.K., 1972, Caste and age variations of the
incidence ofLDH variants in the Bengali Hindus. Humangenetik, 14, 151-154.
Das, S.R., Mukherjee, B.N., Das, S.K., and Malhotra, K.C., 1972, Four types of
genetic variants ofLDH found in India. Human Heredity, 22, 264-270.
Das, S.R., 1973, Tests for incomplete penetrance of the dominant allele in the
heterozygote in the unifactorial autosomal diallelic inheritance. Journal of the
Indian Anthropological Society, 8, 1-12.
Roy M., and Das, S.R., 1973, Blood Group Genetic Survey among the Poundra
Kshatriya Caste of 24 Parganas in West Bengal. Journal of the Indian
Anthropological Society, 8, 1, 83-86.
Das, S.R., Malhotra, K.C., Mukherjee, B.N., and Das, S.K., 1973, Genetic
variants of Lactate dehydrogenase in the Dhangars of Maharashtra, India. Japanese
Journal of Human Genetics, 18,305-308.
360
Das, S.R., Das, S.K, and Dawn, C.S., 1973, Hemoglobin Barts in Bengali Hindu
caste studies in Calcutta. Human Heredity, 23, 381-385.
Das, S.R., Mukherjee, B.N., Das, S.K, Roy, M., and Chhatui, S.S., 1974, Blood
groups, serum proteins, haemoglobins and some serum and red cell enzymes
among Kaoras of24 Parganas in West Bengal (India). Human Heredity, 24, 24-
31.
Das, S.R., Mukherjp,e, B.N., and Das, S.K, 1974, Caste variation in the
distribution of Placental alkaline Phosphatase genes among the Hindus of West
Bengal. Annals of Human Biology, 1,65-71.
Das, S.R., Malhotra, KC., Mukherjee, B.N., and Das, S.K, 1974, LDH and MDH
variants in five castes around Delhi, India. Japanese Journal of Human Genetics,
18,401-404.
Mukherjee, B.N., and Das, S.R., 1974, Distribution of some Polymorphic Enzyme
Group Systems in India. In: Human Population Genetics in India.(edited by
Sanghvi,L.D, Balakrishnan,V., Bhatia, H.M., Sukumaran, P.K, and Undevia,
J.N.) Proceedings of the first Conference of the Indian Society of Human genetics,
held in Bombay February, 14- 16, page, 21-49.0rient Longman Ltd.
Chakraborty, R., Das, S.R., Roy, M., Mukherjee, B.N., and Das, S.K, 1975, The
effect of parity on Placental Weight and birth weight.Interaction with Placental
alkaline phosphatase genotype. Annals of Human Biology, 2, 227-234.
Das, S.R., Roy, M., Paul, A., and Chakraborty, R., 1975, A growth study of
Indian Infants. Non relationship with Placental alkaline phosphatase genotypes.
Human Biology, 47, 219-230.
Chakraborty, R., Roy, M., and Das, S.R., 1975, Proportion of low birth weight
infants in an Indian population and its relationship with maternal age and parity.
Human Heredity, 25,73-74.
Mukherjee, B.N., Das, S.R., 1975, Distribution of some polymorphic enzyme
group systems in India.Proceedings of the International Conference of Indian
Society of Human Genetics Volume, 1.
Das, S.R., Mukherjee, D.P., Bose, L., and Das, A., 1976, A radiological study of
skeletal maturity in hand and wrist area of Bengali Boys and Girls. Memoir no.
38. Anthrpological Survey oflndia, Calcutta.
Das, S.R., 1976, Evaluating Rh chromosome frequencies. Analysis of Lepcha data.
Anthropologist, 23, 1-5.
Mukherjee, B.N., Das, S.K, Malhotra, K.C., and Das, S.R., 1977, Haptoglobin
and acid phosphatase gene distribution in Dhangars of Maharashtra, India.
Journal of Genetics, 63, 39-45.
Hauspie, R.C., Das, S.R., Preece, M.A., and Tanner, J.M., 1980, A longitudinal
study of the growth in height of boys and girls of West Bengal (India) aged six
months to 20 years. Annals of Human Biology, 7, 429-441.
1981 - 1990
Hauspie, R.C., Das,S.R., Preece, M.A, and Tanner, J.M, 1982, Degree of
resemblance of the patterns of growth among siblings in families of West Bengal
(India). Annals of Human Biology, 9,171-174.
Hauspie, R., Das, S.R., Preece, M.A., and Tanner, J.M., 1984, Interrelationships
between various aspects of the growth pattern in weight, and adult sex
361
After 1991
Hauspie, R.C. and Das, S.R., 1995, Short term variations in growth rate in height,
sitting height and biacromial diameter in Bengali Children. In: Essays on
Auxology presented to James Mourilyan Tanner by his former colleagues and
fellows. Edited by R. Hauspie, F. Falkner and G.Lindgren. Castlemead
Publications., 260-268.
Dasgupta, P., and Das, S.R., 1997, A cross-sectional Growth study of trunk and
limb segments of the Bengali Boys of Calcutta. Annals of Human Biology, 24,
363-369.
INDEX
Aborigines, 162 Genetic factors, 123, 127-128,232
Adipose tissue, 91 Genetics, 109,205-206,208-209,212-213,216
Adolescent growth, 92-93, 237-238, 241, 247- Growth and thalassemia, 341
248 Growth charts, 226
Africa (secular trend in -), 161-162, 165 Growth faltering, 251-252, 262-263
Age at growth completion, 21, 26-30 Growth model, 17-23,26,30
Age at peak velocity, 17-18, 21, 27 Growth monitoring, 45-36
American-Japanese parentege, 281 Growth references, 45
Animal short-term growth, 324 Growth retardation, 251-252, 262, 264-265
Appropriate-for-gestational age (AGA), 334 Growth spurt, 238, 243, 245
Australian Aborigines, 109 Growth standards, 46
Growth velocity, 251, 261, 263-265
Basque country, 33
Bengali children, 223 Heritability, 118-119
Bioimpedance analysis, 71 Hormones and short-term growth, 328
Biological parameters, 17-18, 21-22, 26-27, 30, Hungarian National Growth Study, 301
230
Birth weight, 133 India, 223, 237, 341
Blacks (secular trend in -), 161-162 Infectious diseases, 264-265
Body composition, 67-69 Interaction (among genes), 80-81, 85, 87
Body mass index, 67-69,143,171
Body proportions, 142,205-211,213-214,216, Knemometry,321-322
218,314
Body segments (growth of -),234,241,313 Linkage (tree-), 80, 85
Breast feeding, 347-48, 52 Linkage analysis, 80, 82-87
Budapest children, 301 LMS-method, 33-35, 38-40, 225
Logical inductive ability, 186
Calcutta, 91 Longitudinal growth studies, 147, 223
Centile lines, 33-35, 39-41, 225 Longitudinal growth, 251-253, 259, 261, 263
Chaotic growth, 324 Low-birth-weight (LBW), 133, 333-334
Child growth, 252, 263 Low-income countries, 252, 262-264
Cognitive thinking, 179, 181 Lumping and splitting, 80, 84, 87-88
Count-Gompertz model, 17-18,20-23,26,30
Major genes, 80, 86-87
Dental arches, 111-112, 114 Malnutrition, 123, 174-175
Dentofacial morphology, 109 Mastication, 109, 115, 120
Developmental coefficient in neonates, 337 Maya, 210, 212-214, 216
Dietary environment, 127 Mean-constant curve, 17, 23, 25
Dietary habits, 124-125, 127 Menarche, 18,21,23,26-30, 139, 142
Dissimilarities in growth patterns, 234 Mental ability, 179, 181
Distance curve, 18, 226 Meta-analysis, 80, 84
Dual energy X -ray absorptiometry, 68-69 Mexico, 55
Dynamics of growth, 223 Mid-growth spurt, 19,21-22,30
Migration, 214
Early growth studies, 6 Miliary data, 5
Early measuring instruments, 5 Mini-growth spurts, 323
Ethnic differences, 269-270, 274 Mini-knemometry, 322
European studies, 11 Mixed parentage, 281, 288
Multivariate methods, 84
Familial growth data, 232
Family case studies, 123 Nature and nurture, 9
Fat distribution, 91-100 Neonatal growth problems, 333
Feeding recommendations, 45-48, 51-52 Neonates (short-term growth in -), 328
Fels Longitudinal Study, 147 Nepal, 251
Fijian children, 269 Newborn growth, 333
363
364
Non-linear growth model, 22-23, 30, 225 Splitting (Lumping and -),84,87-88
North American studies, 10 Study design, 82-83, 148
Nutritional status, 129, 169-170 Stunting, 251-252, 256, 258-259, 262-265
Obesity, 141-142 Teeth, 111-115, 118
Overweight, 141-142 Thalassemia, 341
The Ancient World, 3
Paediatrics, 67-69 The eighteenth century, 5
Patrilineal families, 126 The Middle Ages, 4
Patrilocal families, 126 The nineteenth century, 7
Peak-height velocity, 233, 240-241, 243, 245 The Renaissance, 4
Peri-urban, 251 Tooth emergence, Ill, 113-1l4
Periodicity in growth, 323 Tree linkage, 80, 85
Physique, 305 Twin studies, 123
Planning a new study, 152
Poland (secular trend in -), 162-165 Undernutrition, 55, 64, 130,240-241,247,249
Population growth studies, 12
Prague children, 314 Velocity curve, 18, 20-22, 226, 238, 245, 226
Preece-Baines model I, 34-35, 40-41, 225 Venezuela, 129
Principal Component Analysis, 91, 93, 95 Verbal ability, 186
Proportions, 142
Psychomotor development, 333 Wasting, 252, 256, 258-259, 262
Public health, 140-141 Weight-for-age, 132
Weight-for-height, 170, 258
Racial classification, 8 Whites (secular trend in -),161-162
Reference curves, 226
Rural area (growth in -), 237-238, 240, 245-247