The Journal of Nutrition
Symposium: A New 21st-Century International Growth Standard for Infants and Young Children
Evaluation of the Feasibility of International
Growth Standards for School-Aged Children
Nancy F. Butte,2* Cutberto Garza,3and Mercedes de Onis4
2USDA/ARS Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX 77030
3Boston College, Chestnut Hill, MA 02467; and4Department of Nutrition for Health and Development, WHO, Geneva, Switzerland
The development of an international growth standard for the screening, surveillance, and monitoring of school-aged
children and adolescents has been motivated by 2 contemporaneous events, the global surge in childhood obesity and the
release of a new international growth standard for infants and preschool children by the WHO. If a prescriptive approach
analogous to that taken by WHO for younger children is to be adopted for school-aged children and adolescents, several
issues need to be addressed regarding the universality of growth potential across populations and the definition of optimal
growth in children and adolescents. A working group of experts in growth and development and representatives from
international organizations concluded that subpopulations exhibit similar patterns of growth when exposed to similar
external conditioners of growth. However, based on available data, we cannot rule out that observed differences in linear
growth across ethnic groups reflect true differences in genetic potential rather than environmental influences. Therefore,
the sampling frame for the development of an international growth standard for children and adolescents must include
multiethnic sampling strategies designed to capture the variation in human growth patterns. A single international growth
standard for school-aged children and adolescents could be developed with careful consideration of the population and
individual selection criteria, study design, sample size, measurements, and statistical modeling of primary growth and
secondary ancillary data. The working group agreed that existing growth references for school-aged children and
adolescents have shortcomings, particularly for assessing obesity, and that appropriate growth standards for these age
groups should be developed for clinical and public health applications.J. Nutr. 137: 153–157, 2007.
The development of an international growth standard for the
screening, surveillance and monitoring of school-aged children
and adolescents has been motivated by 2 contemporaneous
events, the global surge in childhood obesity (1) and the release
of a new international growth standard for infants and pre-
school children by the WHO in collaboration with the United
Nations University (UNU) and other UN agencies, governments,
andnongovernmental organizations (2). Recognition of the limi-
tations of existing growth references that are used for assessing
childhood obesity [e.g., the National Center for Health Statistics
(NCHS)/WHO growth reference (3), the CDC 2000 growth
charts (4), and International Obesity Task Force (IOTF) cutoffs
(5)] has created an urgency and desirability of harmonizing
growth assessment tools conceptually and pragmatically. The
new growth standard for infants and toddlers was developed
from the WHO Multicentre Growth Reference Study (MGRS)
and was released in April, 2006. The MGRS was designed to
describe how children should grow rather than how they grew in
a particular time and place (6,7). In part, the prescriptive ap-
proach upon which the new standard was based required an
expanded definition of ‘‘health,’’ one that went beyond the
absence of overt disease to the adoption of lifestyle practices that
support optimal growth and development. If a prescriptive ap-
proach analogous to that taken by WHO and UNU for younger
children is to be adopted for the development of a growth
standard for school-aged children and adolescents, several issues
would have to be addressed regarding the universality of growth
potential across populations and the characteristics of children
and adolescents most likely to exhibit optimal growth.
1Presented as part of the symposium ‘‘A New 21st-Century International
Growth Standard for Infants and Young Children’’ at the 2006 Experimental
Biology meeting on April 2, 2006, in San Francisco, CA. The symposium was
sponsored by the American Society for Nutrition. The proceedings are published
as a supplement to The Journal of Nutrition. This supplement is the responsibility
of the Guest Editors to whom the Editor of The Journal of Nutrition has delegated
supervision of both technical conformity to the published regulations of The
Journal of Nutrition and general oversight of the scientific merit of each article.
The opinions expressed in this publication are those of the authors and are not
attributable to the sponsors or the publisher, editor, or editorial board of The
Journal of Nutrition. The Guest Editors for the symposium publication are
Mercedes de Onis, World Health Organization, Geneva, Switzerland and
Cutberto Garza, Boston College, Boston, MA.
* To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
0022-3166/07 $8.00 ª 2007 American Society for Nutrition.
This article reviews the feasibility of developing a single international
growth standard with height, weight, and BMI as primary measures for
school-aged children and adolescents. A meeting was convened in
Geneva, Switzerland, on 16–19 January 2006, by the UNU Food and
Nutrition Program, in collaboration with WHO Department of Nutri-
tion for Health and Development and the Food and Agriculture
Organization (FAO) Food and Nutrition Division, to present and discuss
11 position papers addressing theoretical, biological, and pragmatic
issues pertaining to the development of an international growth standard
for school-aged children and adolescents (8–18). In addition to the
authors, representatives from WHO, CDC, UNU, FAO, World Food
Program (WFP), United Nations Children’s Fund (UNICEF), IOTF, and
International Association for Study of Obesity (IASO) attended.
To develop a growth standard for school-aged children and adoles-
cents, 2 fundamental scientific questions needed to be addressed and
were the focus of the 11 commissioned papers and proceedings of the
Geneva meeting. The first question addressed whether it was possible to
develop a single international growth standard for children .5 y of age
that would be representative and useful for the global population, given
the possible genetic differences in growth potential across populations.
The second question was whether a prescriptive approach could be used
to develop a growth standard for school-aged children and adolescents
from either historical and/or prospective growth data.
To address the feasibility of adopting a prescriptive approach to
and adolescents from either historical and/or prospective growth data, it
would be useful to reaffirm the operational difference between growth
a defined population, whereas a standard defines a recommended pattern
of growth that has been associated empirically with specified health
outcomes and the minimization of long-term risks of disease.
Results and Discussion
Limitations of current growth references
The current NCHS/WHO growth reference for children and
adolescents was based on the 1977 NCHS growth charts (19).
Age-sex specific BMI percentiles (20) based on 1971–74
NHANESI data were endorsed for global use by WHO (21).
The NCHS/WHO reference was developed from cross-sectional
data collected from 4 separate samples of children and adoles-
cents surveyed in the U.S. between 1963–1974. The NCHS/
WHO reference may not describe optimal growth given the
extent of its positive skewness in body weight, which is, unfor-
tunately, a drawback shared by other more recent references
such as the CDC 2000 reference and the IOTF cutoffs (22). The
impact of the substantial upward skewness of these 3 references
results in a substantial underestimate of obesity in school-aged
children and adolescents (22–26).
features for the development of a new international growth
standard (21): 1) the sample should represent healthy children
undergoing unconstrained, but not excessive growth, from
several developing and developed countries; 2) secular trends in
growth should be small or absent in the sampled population; 3)
to estimate the more extreme percentiles of weight and height
distributions; and 4) cutoffs for under- and overweight should be
derived in terms of specificity, sensitivity, and positive predictive
values of functional and health-related outcomes.
Interpopulation variation in growth
The technical shortcomings of existing growth references are
addressable, but the fundamental issue challenging the devel-
opment of a single international growth standard is the legit-
imacy of combining subpopulations in light of possible genetic
differences in growth potential. The universality of human
growth was demonstrated for preschool-aged children reared
under favorable nutritional and environmental conditions,
regardless of genetic or ethnic background (27), and elaborated
by Martorell and Habicht (28). The feasibility of developing a
single international growth standard was challenged by Eveleth
and Tanner (29) pointing to differences in achieved height
and growth patterns across subpopulations of children and
To expand on studies tabulated in Worldwide Variation in
Human Growth by Eveleth and Tanner (29), Haas et al. (11)
reviewed the literature since 1988 on interpopulation variation
in achieved height. Growth data of nominally healthy, privi-
leged children across the 5 major geographic regions of Africa,
East Asia, South Asia, West Asia, and Europe, were compiled
and compared with the NCHS/WHO reference. Multiracial
immigrants moving to more advantaged environments were
also included. Major findings from this latest review were as
follows. African children and adolescents of upper socioeco-
nomic status achieve similar heights to the NCHS/WHO ref-
erence medians, although studies are few. African American
boys and girls achieved or exceeded the median values. Mean
heights achieved by East Asian children and adolescents were
below the NCHS/WHO median at all ages between 7 and 18 y,
except for recent values from Beijing (30), and Taiwan (31). In
these studies, heights were similar to the NCHS/WHO reference
until puberty at which time mean heights fell to about the 25th
percentile. Similarly, heights of boys and girls from South and
West Asia tended to follow or were slightly below the NCHS/
WHO until 11–13 y, at which time they fell to ;5 cm below the
reference. The heights of children from central Europe tended to
be 2–4 cm shorter than the NCHS/WHO median, whereas those
from southern and northern countries tended to be similar. At
puberty, mean heights of most European populations approached
the reference median, except for adolescents from Northern
Europe where heights were 4 to 7 cm higher than the reference
at 18 y.
Although the above studies focused on nominally healthy,
privileged children, secular trends in linear growth still may be
occurring in some of the regions. Therefore, Haas et al. (11)
compared the tallest children from various ethnic/geographical
regions that presumably attained their genetic potential in linear
growth. The mean heights of these boys and girls tracked along
the medianof the NCHS/WHO with a mean difference of ;5 cm
between ages of 7 to 13 y. By the age of 15 y, the mean heights of
Mexican American and Japanese adolescents fell to ;5 cm
below and the Dutch means increased to ;5 to 7 cm above the
reference. Mean heights of children living under privileged
conditions worldwide did not vary by .4 cm from 7 y of age
until the initiation of puberty. During adolescence mean heights
in all populations, except those of European origin, were ;5–6
cm (;0.6 SD) below the NCHS/WHO reference median, and
those from Northern Europe exceeded the reference median by
1.0 SD at 18 y of age. It remains to be determined whether these
differences in adolescent linear growth for non-European pop-
ulations represent full attainment or some unrealized gain in
genetic potential. Whether the degree of geographic isolation
and ancestral environmental exposures experienced by some
subpopulations are sufficient to affect the genes that control
linear growth is unknown. If subpopulation differences in height
achieved under optimal environmental conditions persist, ge-
netic differences in growth potential may be responsible.
Genetic determinants of growth
In general, growth parameters, including height and weight, are
highly heritable traits (15). Also, determinants of human
growth, such as the timing and tempo of puberty and other
measures of skeletal and sexual maturation, are largely under
genetic control. Weight, fat mass, and fat distribution are
influenced to a larger extent by environmental factors, although
genetic factors also are significant. Heritability estimates for
growth parameters are lower in nonaffluent populations, prob-
ably due to more pronounced influence of specific nongenetic
factors such as disease and nutrition in those populations.
Limited data are available on cross-population effects of specific
genes or gene variants on growth during childhood and ado-
lescence. Genetic epidemiological studies are needed in different
regions of the world to better explore population differences
in gene frequencies and gene–environmental interactions. Al-
thoughthe fundamentalgeneticunderpinnings ofhumangrowth
are likely to be essentially the same worldwide, frequencies of
allelic gene variants and gene by environmental interactions that
influence growth and maturation might differ across popula-
tions. Their relative influence in different groups, however, re-
In the development of a single international growth standard,
mean growth and normalvariation in growth across populations
must be represented. This should not pose an insurmountable
problem because the largest variance in complex traits such as
weight and height are usually contained within any sufficiently
large sample of children from any given population.
There was a consensus among the working group that
humans follow a similar pattern of growth across ethnic groups
and geographic locations. When exposed to similar external
conditioners of growth, subpopulations exhibit similar patterns
of growth. This was demonstrated years ago for children ,5 y of
age (27) and was more recently confirmed by the WHO MGRS
(32). Although the data for children older than 5 y of age are
more limited, similar growth patterns across subpopulations
were accepted as a general principle by this working group.
Therefore, it was concluded that a single standard can describe
universal human growth patterns. However, based on available
data, it cannot be ruled out that some of the observed differences
in linear growth across ethnic groups reflect true differences in
genetic potential rather than the sole influence of environmental
factors. Therefore, the sampling frame for the development of an
international growth standard for children and adolescents
would have to include multi-ethnic sampling strategies designed
to capture the variation in human growth patterns.
Development of a new international growth standard
In the short term, a cross-sectional growth reference that
approaches a standard for universal use could be constructed
using carefully selected historical data sets that reflect realized
growth potential and good health of school-aged children and
adolescents (9). The reference population should be one that is
stabilized in terms of secular increments in height and weight,
and has not been subjected to discernible external constraints on
growth (dietary deficiencies, infections, etc). But because histor-
ical datasets seldom have detailed subject descriptors, the health
status of the cohort would be unqualified. Furthermore, the
available datasets are not representative of the global popula-
tion, and therefore, such an interim cross-sectional growth
reference should be viewed as provisional.
In the long term, a mixed longitudinal growth standard re-
flective of the multiethnic populations across the regions of the
world could be developed. Using prospective data to develop an
international growth standard, a prescriptive approach is pos-
sible if careful consideration is given to selecting populations or
subgroups that live in communities that support healthy life-
styles and thereby, presumably, optimal growth. Thus, commu-
nities sampled for the development of a growth ‘‘standard’’
would not be representative of nations’ or regions’ populations
but would be uniquely defined on the basis of broadened criteria
for health in a manner analogous to that of the new WHO
growth standard for infants and preschool children (6). Criteria
that specify healthy behaviors at the individual level would be
applied to generate prescriptive-based data. Of great importance
is that the samples selected should be free from obesity as well as
A mixed longitudinal design would produce the most useful
growth data in the shortest period of time (13). The sampling
frame for the development of a prescriptive growth standard
would involve identification of a given number of countries that
is broadly representative of the global community, drawing
samples of children that are subject to inclusion and exclusion
criteria to ensure unconstrained but not excessive growth. The
sample size depends on the complexity of the growth curve;
prior to puberty, the growth patterns for height and weight
are relatively simple, in contrast to the more complex pattern of
the pubertal growth spurt that would require a larger sample
Major environmental influences on growth of children and
adolescents must be considered for the selection of individuals
and populations in the development of an international growth
standard (16,33). Inclusion criteria should encompass adequate
nutrition, lack of significant endemic rates of infection, and
socioeconomic status that does not constrain growth. Low birth
weight, catch-up growth, breast-feeding and early adiposity
rebound can impact growth and/or body composition into
puberty. Exclusion criteria might include low birth weight due to
identifiable pathologies and catch-up growth for individuals and
high altitude and exposure to high levels of environmental
pollution for populations. Populations with minimal evidence of
have been documented in European, European-origin, and Asian
populations where mean heights and weights across generations
have been shown to be greater whereas sexual maturation and
adolescent growth spurts have taken place at progressively
younger ages. The mean secular increase in height in Europe and
North America between 1880 and 1980 was more pronounced
during adolescence because of the tempo effect (2–3 cm per
decade) and less so during childhood (1–2 cm per decade) (29).
In Japan between 1950 and 1980, the secular trend in height was
almost entirely due to the increase in length of the legs. Age
of menarche has been getting earlier during the last century by
;3–4 mo per decade in most European countries. The Japanese
experienced a dramatic decline in the age of menarche between
1950 and 1975; at ;1.0 y per decade in the general population
the age of menarche is as early or earlier than the majority of
European populations. Negative secular trends also have been
seen among populations in Africa, Papua New Guinea, and
Latin America, largely attributable to socioeconomic and polit-
ical deterioration; populations under such psychosocial stress
should be excluded from the sampling frame.
Because biological maturation is closely related to growth,
indicators of biological maturation, including sexual, skeletal,
morphological, and/or dental maturity, must be included in the
data collection for the development of a growth standard (14).
The timing, sequence, and tempo of maturity indicators on
International growth standards for school-aged children and adolescents155
growth must be considered. Skeletal maturation is usually moni-
tored using standardized radiographs, and assessment of matu-
rity is based on changes occurring from initial ossification to
adult morphology of bones of the hand and wrist. Sexual mat-
uration begins with early embryonic differentiation and ends
with full maturity of the sexual organs and fertility. The assess-
ment of sexual maturation is based on secondary sex character-
istics, i.e., breast development, pubic hair, and menarche in girls,
and genital development and pubic hair in boys. Ratings can be
performed by clinical examination or self-examination using
standardized drawings. The mean age at onset of the adolescent
growth spurt occurs between 8.0 and 10.3 y, and the age of peak
height velocity occurs 2 y later (10.8 to 12.2 y) in European and
North American girls. Maturation events occur 2 y later in boys.
Interindividual variation within populations is considerable.
Indicators of sexual maturation, i.e., age at onset, age of peak
height velocity, and skeletal maturity, are recommended indica-
tors of the maturation process.
Direct methods for determining size and structure, including
height, weight, skinfolds, and waist circumference, are well es-
tablished and can be used to monitor linear growth, body mass,
ponderosity, abdominal fat, and fat distribution (18). More
complex body composition methods such as dual X-ray absorp-
tiometry (DXA) and hydrometry would be desirable in view of
mounting evidence of the relation of body fat to cardiovascular
and diabetes risk in children and adolescents.
Measurement of physical activity and physical fitness as
indicators of a healthy lifestyle should be incorporated in the
development of international growth standards for children and
adolescents (17). Physical activity plays an important role in the
regulation of weight, fat mass, and the structural and functional
integrity of bone and skeletal muscle, but probably not height
or the maturation process. Physical fitness changes with age,
growth, and maturation, independent of physical activity. Phy-
sical activity is assessed using questionnaires, interviews, diaries,
direct or indirect (video) observation, film/video and motion
sensors such as pedometers and accelerometers, heart rate moni-
toring, oxygen consumption, and doubly labeled water. Com-
monly used indicators of physical fitness are cardiorespiratory
endurance (endurance shuttle run), function of the lower back
(strength and flexibility), and many health-related fitness tests.
Although data for children and adolescents are limited, they
suggest a relation, although moderate, of physical activity and
fitness to a favorable risk profile.
In conclusion, the working group agreed that the NCHS/
WHO growth reference for school-aged children and adoles-
cents, the CDC 2000 growth charts, and the IOTF cutoffs, all
have shortcomings, and that a more appropriate growth stan-
dard for clinical and public health applications for these age
groups should be developed. An international growth standard
for school-aged children and adolescents could be constructed
with careful consideration of the population and individual
selection criteria, study design, sample size, measurements, and
statistical modeling of primary growth and secondary ancillary
1. Lobstein T, Baur L, Uauy R. Obesity in children and young people: a
crisis in public health. Obes Rev. 2004;5:4–104.
WHO Multicentre Growth Reference Study Group. WHO Child
Growth Standards: Length/height-for-age, weight-for-age, weight-for-
length, weight-for-height and body mass index-for-age: methods and
development. Geneva: World Health Organization, 2006.
3. Wang Y, Wang JQ. A comparison of international references for the
assessment of child and adolescent overweight and obesity in different
populations. Eur J Clin Nutr. 2002;56:973–82.
Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Strawn LM, Flegal KM,
Mei Z, Wei R, Curtin LR, Roche AF, Johnson CL. 2000 CDC growth
charts for the United States: methods and development. Vital Health
Stat 11. 2002;246:1–190.
Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard
definition for child overweight and obesity worldwide: international
survey. BMJ. 2000;320:1240–3.
de Onis M, Garza C, Victora CG, Onyango AW, Frongillo EA, Martines
J, WHO Multicentre Growth Reference Study Group. The WHO Mul-
ticentre Growth Reference Study: planning, study design, and method-
ology. Food Nutr Bull. 2004;25:S15–26.
Garza C, de Onis M, WHO Multicentre Growth Reference Study
Group. Rationale for developing a new international growth reference.
Food Nutr Bull. 2004;25:S5–S14.
Wang Y, Moreno LA, Caballero B, Cole TJ. Limitations of the current
World Health Organization growth references for children and adoles-
cents. Food Nutr Bull. 2006;27(suppl):S175–88.
Seidell JC, Doak CM, de Munter JSL, Kuijper LDJ, Zonneveld C. Cross-
sectional growth references and implications for the development of an
international growth reference for children and adolescents. Food Nutr
10. Himes JH. Long-term longitudinal studies and implications for the de-
velopment of an international growth reference for children and
adolescents. Food Nutr Bull. 2006;27(suppl):S199–211.
11. Haas JD, Campirano F. Interpopulation variation in height from 7 to 18
years of age. Food Nutr Bull. 2006;27(suppl):S212–23.
12. Pelletier D. Theoretical considerations related to cutoff points. Food
Nutr Bull. 2006;27(suppl):S224–36.
13. Cole TJ. Development of an international growth standard for preado-
lescent and adolescent children. Statistical considerations. Food Nutr
14. Beunen GP, Rogol AD, Malina RM. Indicators of biological maturation
and secular changes in biological maturation. Food Nutr Bull.
15. Thomis MM, Towne B. Genetic determinants of pre-pubertal and
pubertal growth and development. Food Nutr Bull. 2006;27(suppl):
16. Ulijaszek SJ. The International Growth Reference for Children and Ado-
lescents project. Environmental influences on preadolescent and ado-
lescent growth in weight and height. Food Nutr Bull. 2006;27(suppl):
17. Malina RM, Katzmarzyk PT. Physical activity and fitness in an
international growth standard. Food Nutr Bull. 27(suppl):S295–313.
18. Lohman TG, Going SB. Body composition assessment for development
of International Growth Standards for Pre-Adolescent and Adolescent
Children. Food Nutr Bull. 2006;27(suppl):S314–25.
19. Hamill PV, Drizd TA, Johnson CL. NCHS growth curves for children
birth–18 years. Vital Health Stat. 1977;11:i-74.
20. Must A, Dallal GE, Dietz WH. Reference data for obesity: 85th and
95th percentiles of body mass index (wt/ht2) and triceps skinfold thick-
ness. Am J Clin Nutr. 1991;53:839–46.
21. World Health Organization. Physical status: the use and interpre-
tation of anthropometry. Report of a WHO Expert Committee.
Technical Report Series No. 854. Geneva: World Health Organization,
22. de Onis M. The use of anthropometry in the prevention of childhood
overweight and obesity. Int J Obes Relat Metab Disord. 2004;28:
23. de Onis M, Dasgupta P, Saha S, Sengupta D, Blossner M. The National
Center for Health Statistics reference and the growth of Indian ado-
lescent boys. Am J Clin Nutr. 2001;74:248–53.
24. Kain J, Uauy R, Vio F, Albala C. Trends is overweight and obesity
prevalence in Chilean children: comparison of three definitions. Eur J
Clin Nutr. 2002;56:200–4.
25. Vignerova J, Lhotska L, Blaha P. Proposed standard definition for child
overweight and obesity. Cent Eur J Public Health. 2001;9:145–6.
26. Fu WP, Lee HC, Ng CJ, Tay YK, Kau CY, Seow CJ, Siak JK, Hong CY.
Screening for childhood obesity: international versus population-specific
definitions. Which is more appropriate? Int J Obes Relat Metab Disord.
27. Habicht J-P, Martorell R, Yarbrough C, Malina RM, Klein RE. Height
and weight standards for preschool children. How relevant are ethnic
differences in growth potential? Lancet. 1974;1:611–4.
28. Martorell R, Habicht J-P. Growth in early childhood in developing
countries. In: Falkner F, Tanner JM, editors. Human growth: a compre-
hensive treatise. 2nd ed. New York: Plenum Press; 1986. p. 241–62.
29. Eveleth P, Tanner JM. Worldwide variation in human growth. 2nd ed.
Cambridge: Cambridge University Press; 1990.
30. Li H, Leung SS, Lam PK, Zhang X, Chen XX, Wang SL. Height and
weight percentile curves of Beijing children and adolescents 0–18 years,
1995. Ann Hum Biol. 1999;26:457–71.
31. Chen JY, Chang H, Pan WH. A modified locally weighted method
for developing reference standards for height, weight, and body mass
index of boys and girls aged 4–18 in Taiwan. Hum Biol. 2003;75:
32. WHO Multicentre Growth Reference Study Group. Assessment of
differences in linear growth among populations in the WHO Multi-
centre Growth Reference Study. Acta Paediatr Suppl. 2006;450:
33. Bogin B. Cambridge studies in biological and evolutionary anthropol-
ogy. Patterns of human growth. 2nd ed. Cambridge: Cambridge Univer-
sity Press; 1999.
International growth standards for school-aged children and adolescents157