ArticlePDF Available

De Onis M, Onyango AW, Borghi E, Siyam A, Nishida C, Siekmann J. Development of a WHO growth reference for school-aged children and adolescents. Bull World Health Organ. 2007;85:660-7

Authors:

Abstract and Figures

To construct growth curves for school-aged children and adolescents that accord with the WHO Child Growth Standards for preschool children and the body mass index (BMI) cut-offs for adults. Data from the 1977 National Center for Health Statistics (NCHS)/WHO growth reference (1-24 years) were merged with data from the under-fives growth standards' cross-sectional sample (18-71 months) to smooth the transition between the two samples. State-of-the-art statistical methods used to construct the WHO Child Growth Standards (0-5 years), i.e. the Box-Cox power exponential (BCPE) method with appropriate diagnostic tools for the selection of best models, were applied to this combined sample. The merged data sets resulted in a smooth transition at 5 years for height-for-age, weight-for-age and BMI-for-age. For BMI-for-age across all centiles the magnitude of the difference between the two curves at age 5 years is mostly 0.0 kg/m(2) to 0.1 kg/m(2). At 19 years, the new BMI values at +1 standard deviation (SD) are 25.4 kg/m(2) for boys and 25.0 kg/m(2) for girls. These values are equivalent to the overweight cut-off for adults (> or = 25.0 kg/m(2)). Similarly, the +2 SD value (29.7 kg/m(2) for both sexes) compares closely with the cut-off for obesity (> or = 30.0 kg/m(2)). The new curves are closely aligned with the WHO Child Growth Standards at 5 years, and the recommended adult cut-offs for overweight and obesity at 19 years. They fill the gap in growth curves and provide an appropriate reference for the 5 to 19 years age group.
Content may be subject to copyright.
660
Bulletin of the World Health Organization | September 2007, 85 (9)
Objective To construct growth curves for school-aged children and adolescents that accord with the WHO Child Growth Standards
for preschool children and the body mass index (BMI) cut-offs for adults.
Methods Data from the 1977 National Center for Health Statistics (NCHS)/WHO growth reference (1–24 years) were merged
with data from the under-fives growth standards’ cross-sectional sample (18–71 months) to smooth the transition between the
two samples. State-of-the-art statistical methods used to construct the WHO Child Growth Standards (0–5 years), i.e. the Box-Cox
power exponential (BCPE) method with appropriate diagnostic tools for the selection of best models, were applied to this combined
sample.
Findings The merged data sets resulted in a smooth transition at 5 years for height-for-age, weight-for-age and BMI-for-age.
For BMI-for-age across all centiles the magnitude of the difference between the two curves at age 5 years is mostly 0.0 kg/m² to
0.1 kg/m². At 19 years, the new BMI values at +1 standard deviation (SD) are 25.4 kg/m² for boys and 25.0 kg/m² for girls. These
values are equivalent to the overweight cut-off for adults (> 25.0 kg/m²). Similarly, the +2 SD value (29.7 kg/m² for both sexes)
compares closely with the cut-off for obesity (> 30.0 kg/m²).
Conclusion The new curves are closely aligned with the WHO Child Growth Standards at 5 years, and the recommended adult
cut-offs for overweight and obesity at 19 years. They fill the gap in growth curves and provide an appropriate reference for the 5
to 19 years age group.
Bulletin of the World Health Organization 2007;85:660–667.
Une traduction en français de ce résumé figure à la fin de l’article. Al final del artículo se facilita una traducción al español.
Development of a WHO growth reference for school-aged
children and adolescents
Mercedes de Onis,
a
Adelheid W Onyango,
a
Elaine Borghi,
a
Amani Siyam,
a
Chizuru Nishida
a
& Jonathan Siekmann
a

Research
a
Department of Nutrition, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland. Correspondence to Mercedes de Onis
(e-mail: deonism@who.int).
doi: 10.2471/BLT.07.043497
(
Submitted: 25 April 2007 – Final revised version received: 12 July 2007 – Accepted: 15 July 2007
)
Introduction
e need to develop an appropriate sin-
gle growth reference for the screening,
surveillance and monitoring of school-
aged children and adolescents has been
stirred by two contemporary events: the
increasing public health concern over
childhood obesity
1
and the April 2006
release of the WHO Child Growth
Standards for preschool children based
on a prescriptive approach.
2
As countries
proceed with the implementation of
growth standards for children under 5
years of age, the gap across all centiles
between these standards and existing
growth references for older children has
become a matter of great concern. It is
now widely accepted that using descrip-
tive samples of populations that reflect
a secular trend towards overweight and
obesity to construct growth references
results inadvertently in an undesirable
upward skewness leading to an underes-
timation of overweight and obesity, and
an overestimation of undernutrition.
3
e reference previously recom-
mended by WHO for children above 5
years of age, i.e. the National Center for
Health Statistics (NCHS)/WHO inter-
national growth reference,
4
has several
drawbacks.
5
In particular, the body mass
index-for-age reference, developed in
1991,
6
only starts at 9 years of age,
groups data annually and covers a lim-
ited percentile range. Many countries
pointed to the need to have body mass
index (BMI) curves that start at 5 years
and permit unrestricted calculation of
percentile and z-score curves on a con-
tinuous age scale from 5 to 19 years.
e need to harmonize growth as-
sessment tools conceptually and prag-
matically prompted an expert group
meeting in January 2006 to evaluate the
feasibility of developing a single inter-
national growth reference for school-
aged children and adolescents.
7,8
e
experts agreed that appropriate growth
references for these age groups should
be developed for clinical and public
health applications. ey also agreed
that a multicentre study, similar to the
one that led to the development of the
WHO Child Growth Standards for 0 to
5 years, would not be feasible for older
children, as it would not be possible to
control the dynamics of their environ-
ment. erefore, as an alternative, the ex-
perts suggested that a growth reference
be constructed for this age group using
existing historical data and discussed the
criteria for selecting the data sets.
WHO subsequently initiated a pro-
cess to identify existing data sets from
various countries. is process resulted
in an initial identification of 115 candi-
date data sets from 45 countries, which
were narrowed down to 34 data sets
661
Bulletin of the World Health Organization | September 2007, 85 (9)
Research
Growth curves for school-aged children and adolescents
Mercedes de Onis et al.
from 22 countries that met the inclusion
criteria developed by the expert group.
However, after further review, even these
most promising studies showed great
heterogeneity in methods and data qual-
ity, sample size, age categories, socioeco-
nomic status of participating children
and various other factors critical to
growth curve construction. erefore,
it was unlikely that a growth reference
constructed from these heterogeneous
data sets would agree with the WHO
Child Growth Standards at 5 years of
age for the different anthropometric
indicators needed (i.e. height-for-age,
weight-for-age and BMI-for-age).
In consequence, WHO proceeded
to reconstruct the 1977 NCHS/WHO
growth reference from 5 to 19 years,
using the original sample (a non-obese
sample with expected heights), supple-
mented with data from the WHO Child
Growth Standards (to facilitate a smooth
transition at 5 years), and applying the
state-of-the-art statistical methods
9,10
used to develop standards for preschool
children, that is, the Box-Cox power
exponential (BCPE) method with ap-
propriate diagnostic tools for the selec-
tion of best models.
e purposes of this paper are to re-
port the methods used to reconstruct the
1977 NCHS/WHO growth reference,
to compare the resulting new curves (the
2007 WHO reference) with the 1977
NCHS/WHO charts, and to describe
the transition at 5 years of age from the
WHO standards for under-fives to these
new curves for school-aged children and
adolescents.
Methods
Sample description
e core sample used for the reconstruc-
tion of the reference for school-aged
children and adolescents (5–19 years)
was the same as that used for the con-
struction of the original NCHS charts,
pooling three data sets.
11
e first and
second data sets were from the Health
Examination Survey (HES) Cycle II
(6–11 years) and Cycle III (12–17 years).
e third data set was from the Health
and Nutrition Examination Survey
(HANES) Cycle I (birth to 74 years),
from which only data from the 1 to 24
years age range were used. Given the
similarity of the three data sets,
11
the data
were merged without adjustments.
e total sample size was 22 917
(11 410 boys, 11 507 girls). For the in-
dicator height-for-age, 8 boys (0.07%),
including an 18 month-old with length
51.6 cm, and 14 girls (0.12%) had
outlier height measurements that were
set to missing in the data set. For the
weight-based indicators (i.e. weight-
for-age and BMI-for-age), the same
cleaning approach used for the con-
struction of the WHO Child Growth
Standards (cross-sectional component)
was applied to avoid the influence of
unhealthy weights-for-height.
10
As a
result, 321 observations for boys (2.8%)
and 356 observations for girls (3.0%)
were excluded.
A smooth transition from the
WHO Child Growth Standards (0–5
years) to the reference curves beyond 5
years was provided by merging data from
the growth standards’ cross-sectional
sample (18–71 months) with the NCHS
nal sample before fitting the new
growth curves. e growth curves for
ages 5 to 19 years were thus constructed
using data from 18 months to 24 years.
e final sample used for tting the
growth curves included 30 907 observa-
tions (15 537 boys, 15 370 girls) for the
height-for-age curves, 30 100 observa-
tions (15 136 boys, 14 964 girls) for
the weight-for-age curves, and 30 018
observations (15 103 boys, 14 915 girls)
for the BMI-for-age curves.
Statistical methods
As the goal was to develop growth
curves for school-aged children and
Fig. 1. Comparison between the 1977 and 2007 height-for-age
z
-score curves
– boys
adolescents that accord with the WHO
Child Growth Standards for preschool
children, we reapplied the state-of-the-
art statistical methods used to construct
the growth standards for children under
5 years of age.
10
e development of
the standards for under-fives followed
a methodical process that involved: (a)
detailed examination of existing meth-
ods, including types of distributions and
smoothing techniques; (b) selection of
a software package flexible enough to
allow comparative testing of alternative
methods and the actual generation of the
curves; and (c) systematic application of
the selected approach to the data to gen-
erate models that best fitted the data.
9
e BCPE method,
12
with curve
smoothing by cubic splines, was used
to construct the curves. is method
accommodates various kinds of dis-
tributions, from normal to skewed or
kurtotic. After the model was fitted us-
ing the whole age range (18 months to
24 years), the curves were truncated to
cover the required age range (i.e. 5–19
years for height-for-age and BMI-for-
age, and 5–10 years for weight-for-age),
thus avoiding the right- and left-edge
effects.
9
e specications of the BCPE
models that provided the best fit to gen-
erate the growth curves were:
For height-for-age:
BCPE(l = 1, df(m) = 12, df(s) = 4, n = 1,
t = 2) for boys
BCPE(l = 0.85, df(m) = 10, df(s) = 4,
n = 1, t = 2) for girls.
Height (cm)
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
100
110
120
130
140
150
160
170
180
190
200
0
-1
-2
-3
1
2
3
Age (years)
2007
1977
Height (cm)
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
100
110
120
130
140
150
160
170
180
190
200
0
-1
-2
-3
1
2
3
Age (years)
2007
1977
662
Bulletin of the World Health Organization | September 2007, 85 (9)
Research
Growth curves for school-aged children and adolescents
Mercedes de Onis et al.
Fig. 2. Comparison between the 1977 and 2007 height-for-age
z
-score curves
– girls
Fig. 3. Comparison between the 1977 and 2007 weight-for-age
z
-score curves
– boys
For weight-for-age:
BCPE(l = 1.4, df(m) = 10, df(s) = 8,
df(n) = 5, t = 2) for boys
BCPE(l = 1.3, df(m) = 10, df(s) = 3,
df(n) = 3, t = 2) for girls.
For BMI-for-age:
BCPE(l = 0.8, df(m) = 8, df(s) = 4,
df(n) = 4, t = 2) for boys
BCPE(l = 1, df(m) = 8, df(s) = 3,
df(n) = 4, t = 2) for girls.
Where l is the power of the transfor-
mation applied to age before fitting the
model; df(m) is the degrees of freedom
for the cubic splines fitting the median
(m); df(s) the degrees of freedom for
the cubic splines fitting the coefficient
of variation (s); df(n) the degrees of
freedom for the cubic splines fitting the
Box-Cox transformation power (n) (for
height-for-age fixed n = 1); and t is the
parameter related to the kurtosis (in all
three cases fixed t = 2).
e selected models for boys and
girls ultimately simplify to the LMS
method,
13
since it was not necessary to
model the parameter related to kurtosis.
For height-for-age, the data follow the
standard normal distribution, so it was
not necessary to model either the param-
eter related to skewness or to kurtosis.
Results
Percentile and z-score curves and tables
were generated ranging from the 1st to
the 99th percentile and from the –3 to
for boys (Fig. 1) than it is for girls (Fig.
2), especially at the upper end of the
age range (15–18 years; 18 years is the
maximum age limit of the 1977 curves).
Dierences in boys’ attained height
z-scores (1977 versus 2007 curves) at 5
years are negligible, ranging from 0.1 cm
in the curves below the median to 0.3 cm
at +2 and +3 SD (Fig. 1). e two sets
of curves follow more variable patterns
in both shape and the spread of attained
heights as they advance from age 10 years
to the end of the age range. For example,
at 18 years, the distribution of heights
from –3 to +3 SD is tighter by 5 cm in
the 1977 curves compared with those
for 2007. Between –3 SD and the me-
dian, the 1977 curves are higher by 3.3
cm, 2.4 cm, 1.5 cm and 0.7 cm, respec-
tively. Conversely, the 1977 curves above
the median are lower than corresponding
2007 curves by 0.2 cm (+1 SD), 1.1 cm
(+2 SD) and 2.0 cm (+3 SD).
Although the disparity at 5 years
between the two sets of girls’ curves
(Fig. 2) is greater than that observed for
boys, ranging between 0.2 cm (–3 SD)
and 1.7 cm (+3 SD), the curve shapes in
later years follow more comparable pat-
terns and culminate in a more similar
distribution of attained height z-scores
between 15 and 18 years of age. As ob-
served for boys, the negative SDs and
median of the 1977 set at 18 years are
higher than their equivalent 2007 curves
by 2.6 cm (–3 SD), 2.0 cm (–2 SD), 1.2
cm (–1 SD) and 0.6 cm (median). e
the +3 standard deviation (SD). e full
set of clinical charts and tables displayed
by sex and age (years and months), per-
centile and z-score values and related in-
formation (e.g. LMS values) are provided
on the WHO web site (http://www.who.
int/growthref/).
Sex-specic comparisons of the
1977 NCHS/WHO reference and the
newly reconstructed curves are presented
in the figures for height-for-age, weight-
for-age and BMI-for-age, respectively.
Height-for-age
e difference in shape between the
1977 and 2007 curves is more evident
Height (cm)
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
100
110
120
130
140
150
160
170
180
0
-1
-2
-3
1
2
3
Age (years)
2007
1977
Height (cm)
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
100
110
120
130
140
150
160
170
180
0
-1
-2
-3
1
2
3
Age (years)
2007
1977
Weight (kg)
5 6 7 8 9 10
20
30
40
50
0
-1
-2
-3
1
2
3
Age (years)
2007
1977
Weight (kg)
5 6 7 8 9 10
20
30
40
50
0
-1
-2
-3
1
2
3
Age (years)
2007
1977
663
Bulletin of the World Health Organization | September 2007, 85 (9)
Research
Growth curves for school-aged children and adolescents
Mercedes de Onis et al.
+1 SD curves overlap at 18 years, and
in reverse pattern to the negative SDs,
the 1977 curves are lower by 0.7 cm (+2
SD) and 1.3 cm (+3 SD).
Weight-for-age
In the lower half of the weight-for-age
distribution, the largest difference be-
tween the 1977 and 2007 boyscurves
(Fig. 3) is at 10 years of age, where the
2007 curves are higher by 2.9 kg (–3
SD) and 1.1 kg (–2 SD). In the upper
half of the distribution, the largest dis-
parities between the +1 SD and +2 SD
curves are also at age 10 years, but in this
case the 1977 curves are higher by 1.7 kg
and 1.0 kg. e +3 SD curves present
sizeable differences, with the 1977 curve
being consistently lower throughout the
age range (from 1.6 kg at 5 years to 3.1
kg at 10 years). Girls present similar pat-
terns to those observed for boys (Fig. 4).
At the lower bound, disparities are larger
for girls than they are for boys. For girls
at 10 years, the 2007 curves are higher
by 3.7 kg (–3 SD) and 1.4 kg (–2 SD).
At the upper bound, the largest disparity
for the +3 SD curves is at 5 years, where
the 2007 curve is 3.1 kg above the 1977
curve, but the difference decreases to
1.7 kg at 10 years. e +2 SD curves
cross between 8 and 9 years. At 5 years,
the 2007 curve is higher by 1.3 kg and,
at 10 years, it is lower than the 1977
curve by 2.3 kg.
BMI-for-age
Fig. 5 (boys) and Fig. 6 (girls) show the
reference data for BMI-for-age devel-
oped in 1991 that WHO has to date rec-
ommended for ages 9 to 24 years
6
and
how they compare with corresponding
centiles of the newly constructed curves
in the age period where the two sets
overlap (9–19 years). e 5th, 15th and
50th percentiles for boys (Fig. 5) start at
9 years with small differences (0.1 kg/m²
and 0.2 kg/m²) between the 1991 refer-
ence values and the 2007 curves. e
two sets then track closely and cross over
at about 17 years, so that by 19 years the
2007 percentiles are 0.3 kg/m² or 0.4
kg/m² higher than the 1991 reference
values. e 85th percentile of the 1991
reference originates at 0.9 kg/m² above
its 2007 equivalent and tracks above it
to end at 0.8 kg/m² higher at 19 years.
For the 95th percentile, the 1991 refer-
ence starts at 2.0 kg/m² higher and veers
upwards, terminating 2.6 units above
Fig. 4. Comparison between the 1977 and 2007 weight-for-age
z
-score curves
– girls
the 2007 curve at 19 years. e patterns
observed in the boys’ curves are also
evident among girls (Fig. 6), except that
the crossover of the 5th, 15th and 50th
percentiles occurs at 13 years, and differ-
ences in the 50th and 95th percentiles
are slightly larger than corresponding
differences in the boys percentiles. A
wiggly pattern is noticeable in the 1991
reference values, particularly in the 50th,
85th and 95th percentiles.
At 19 years of age, the 2007 BMI
values at +1 SD are 25.4 kg/m² for boys
and 25.0 kg/m² for girls, while the
+2 SD values are 29.7 kg/m² for both
sexes.
Transition to the 2007 WHO
reference at 5 years
A main objective for reconstructing the
1977 NCHS/WHO reference was to
provide a smooth transition from the
WHO standard curves for under-fives
to the reference curves for older children.
Table 1 presents values at 5 years for the
various indicators by sex of the 1977
and 2007 references for school-aged
children and adolescents, and the WHO
standards for under-fives.
Disparities between the 1977 refer-
ence and the WHO height-for-age and
weight-for-age standards for girls at 5
years were larger than those observed
in corresponding boys’ curves. For
example, the dierences in the boys’
height-for-age curves were at most 0.2
cm, in contrast to the girlscurves that
were disparate by as much as 1.7 cm and
2.1 cm at +2 and +3 SD, respectively.
For weight-for-age, differences between
the 1977 reference and the WHO stan-
dards at +3 SD were 2.0 kg for boys and
3.5 kg for girls. Since no NCHS-based
reference values for BMI were available
for ages below 9 years, the table presents
comparative values only for the 2007
reconstructed reference and the WHO
standards at 5 years of age.
e reconstruction resulted in
curves that are closely aligned to cor-
responding WHO standards at the
junction (5 years). For height-for-age
boys, the three negative SDs are only
0.1 cm apart, the median and +1 SD
curves differ by 0.3 cm, and disparities
at +2 SD and +3 SD are 0.4 cm and
0.5 cm, respectively. For girls, the differ-
ences between the two sets of curves are
0.3 cm or 0.4 cm through the full range
of z-scores. For weight-for-age, where
differences between the 1977 reference
and the WHO standards at 5 years were
considerable, the reconstruction sub-
stantially reduced differences in the final
curves. e boys’ medians are equal,
while their negative z-scores differ by 0.1
kg or 0.2 kg, and the positive z-scores
by 0.1 kg (+1 SD), 0.3 kg (+2 SD) and
0.4 kg (+3 SD). Residual differences in
the two sets of curves for girls are in a
range similar to those in the boyscurves,
which is between 0.0 kg and 0.4 kg.
e merger of the under-ves
growth standardsdata (18–71 months)
with the NCHS core sample to fit the
2007 curves for school-aged children
and adolescents resulted in a very
smooth transition between the WHO
Child Growth Standards and the newly
Weight (kg)
5 6 7 8 9 10
10
20
30
40
50
60
0
-1
-2
-3
1
2
3
Age (years)
2007
1977
Weight (kg)
5 6 7 8 9 10
10
20
30
40
50
60
0
-1
-2
-3
1
2
3
Age (years)
2007
1977
664
Bulletin of the World Health Organization | September 2007, 85 (9)
Research
Growth curves for school-aged children and adolescents
Mercedes de Onis et al.
Fig. 5. Comparison between the 1991 and 2007 body mass index-for-age
percentile curves – boys
constructed references for BMI-for-age.
For both boys and girls, differences be-
tween the two curve sets at 5 years are
mostly 0.0 kg/ or 0.1 kg/, and
never more than 0.2 kg/m².
Discussion
e need for a widely applicable growth
reference for older children and ado-
lescents was increasingly recognized
by countries attempting to assess the
magnitude of the growing public health
problem of childhood obesity. is
need was reaffirmed by the release of
the under-five growth standards. e
reconstruction presented in this paper
has resulted in growth curves that are
closely aligned to the WHO Child
Growth Standards at 5 years and as such
are a suitable complementary reference
for use in school-aged child and adoles-
cent health programmes. e various
clinical charts and tables provided on
the Internet will allow for the practical
application of the reference.
e approach used in constructing
the 2007 WHO reference addressed the
limitations of the 1977 NCHS curves
recognized by the 1993 expert commit-
tee
4
that recommended their provisional
use for older children. e height-for-
age median curves of the 1977 and 2007
references overlap almost completely
with only a slight difference in shape,
which is probably due to the different
modelling techniques used. For the
1977 NCHS/WHO curves, age-specific
standard deviations from the median
were derived from the observed disper-
sion of six percentile curves (5th, 10th,
25th, 75th, 90th and 95th) and then
smoothed by a combination of polyno-
mial regression and cubic splining tech-
niques.
14
In the 2007 reconstruction, age
was modelled as a continuous variable,
and the curves were fitted simultane-
ously and smoothed throughout the age
range using cubic splines. Furthermore,
edge effects were avoided by construct-
ing the 2007 curves with data that
extended beyond the lower and upper
age bounds of the final reference curves.
e latter may explain why the 1977
NCHS/WHO curves have pronounced
wiggly shapes towards the upper age
limit of the reference compared with the
2007 curves.
When compared to the 1977
NCHS/WHO curves, the differences in
the newly reconstructed weight-for-age
curves are significant in all centiles apart
from the median and the –1 SD curves,
reflecting the important difference in
curve construction methodology. e
fact that the median curves of the two
references overlap almost completely is
reassuring in that the two samples used
for fitting the models are the same within
the healthy range (i.e. middle range of
the distribution). e methodology
available at the time of constructing the
1977 curves was limited in its ability to
model skewed data.
14
Fixing a higher
standard deviation distance between the
curves above the median and a lower
one for the curves below, as was done,
partially accounted for the skewness
in the weight data but failed to model
the progressively increasing distances
between the SD curves from the lower
to the upper tails of the weight-for-age
distribution. To t the skewed data
adequately, the LMS method (used in
the construction of the 2007 curves and
other recently developed weight-based
references) fits a Box-Cox normal dis-
tribution, which follows the empirical
data closely.
15–17
e reference data for BMI-for-age
recommended by WHO are limited in
that they begin only at 9 years of age
and cover a restricted distribution range
(5th–95th percentiles). e empirical
Fig. 6. Comparison between the 1991 and 2007 body mass index-for-age
percentile curves – girls
665
Bulletin of the World Health Organization | September 2007, 85 (9)
Research
Growth curves for school-aged children and adolescents
Mercedes de Onis et al.
reference values were estimated us-
ing data that were grouped by age in
years, and then smoothed using locally
weighted regression.
6
e 2007 recon-
struction permits the extension of the
BMI reference to 5 years, where the
curves match WHO under-five curves
almost perfectly. Furthermore, at 19
years of age, the 2007 BMI values for
both sexes at +1 SD (25.4 kg/m² for
boys and 25.0 kg/m² for girls) are
equivalent to the overweight cut-o
Table 1. Reference values for height-for-age, weight-for-age and body mass index-for-age at 5 years by sex for the 1977 and
2007 references, and the WHO Child Growth Standards
1977 reference 2007 reference WHO standards
a
1977 reference 2007 reference WHO standards
a
Z
-scores
Boys Girls
Height-for-age (cm)
3 SD 96.1 96.0 96.1 95.1 94.9 95.2
2 SD 100.7 100.6 100.7 99.5 99.6 99.9
1 SD 105.3 105.2 105.3 104.0 104.3 104.7
Median 109.9 109.7 110.0 108.4 109.1 109.4
+1 SD 114.5 114.3 114.6 112.8 113.8 114.2
+2 SD 119.1 118.8 119.2 117.2 118.6 118.9
+3 SD 123.7 123.4 123.9 121.6 123.3 123.7
Weight-for-age (kg)
3 SD 12.3 12.6 12.4 11.9 12.2 12.1
2 SD 14.4 14.2 14.1 13.8 13.8 13.7
1 SD 16.6 16.1 16.0 15.7 15.8 15.8
Median 18.7 18.3 18.3 17.7 18.1 18.2
+1 SD 21.1 20.9 21.0 20.4 21.0 21.2
+2 SD 23.5 23.9 24.2 23.2 24.5 24.9
+3 SD 25.9 27.5 27.9 26.0 29.1 29.5
Body mass index-for-age (kg/m²)
b
3 SD 12.1 12.0 11.8 11.6
2 SD 13.0 12.9 12.8 12.7
1 SD 14.1 14.0 13.9 13.9
Median 15.3 15.2 15.2 15.3
+1 SD 16.6 16.6 16.9 16.9
+2 SD 18.2 18.3 18.8 18.8
+3 SD 20.1 20.3 21.3 21.1
a
WHO Child Growth Standards for 0–5 years of age.
2,10
b
For BMI, the 1991 reference data start at 9 years of age.
4
used for adults (> 25.0 kg/m²), while
the +2 SD value (29.7 kg/m² for both
sexes) compares closely with the cut-off
for obesity (> 30.0 kg/m²).
18
e 2007 height-for-age and BMI-
for-age charts extend to 19 years, which
is the upper age limit of adolescence as
defined by WHO.
19
e weight-for-age
charts extend to 10 years for the ben-
efit of countries that routinely measure
only weight and would like to monitor
growth throughout childhood. Weight-
for-age is inadequate for monitoring
growth beyond childhood due to its
inability to distinguish between relative
height and body mass, hence the provi-
sion here of BMI-for-age to complement
height-for-age in the assessment of thin-
ness (low BMI-for-age), overweight and
obesity (high BMI-for-age) and stunting
(low height-for-age) in school-aged chil-
dren and adolescents. O
Competing interests: None declared.
Résumé
Mise au point d’une référence de croissance pour les enfants d’âge scolaire et les adolescents
Objectif Construire des courbes de croissance pour les enfants
d’âge scolaire et les adolescents concordant avec la Norme OMS
de croissance de l’enfant pour les enfants d’âge préscolaire et
avec les points de coupure pour l’indice de masse corporelle (IMC)
s’appliquant aux adultes.
Méthodes Les données de référence NCHS/OMS pour la
croissance (de 1 à 24 ans) de 1977 ont été regroupées avec
celles de l’échantillon transversal d’enfants de moins de 5 ans
(18 à 71 mois) utilisé pour la norme de croissance de manière
à lisser la transition entre les deux échantillons. Les méthodes
666
Bulletin of the World Health Organization | September 2007, 85 (9)
Research
Growth curves for school-aged children and adolescents
Mercedes de Onis et al.
Resumen
Elaboración de valores de referencia de la OMS para el crecimiento de escolares y adolescentes
Objetivo Elaborar curvas de crecimiento para escolares y
adolescentes que concuerden con los Patrones de Crecimiento
Infantil de la OMS para preescolares y los valores de corte del
índice de masa corporal (IMC) para adultos.
Métodos Se fusionaron los datos del patrón internacional de
crecimiento del
National Center for Health Statistics
/OMS de
1977 (1–24 años) con los datos de la muestra transversal de los
patrones de crecimiento para menores de 5 años (18–71 meses),
con el fin de suavizar la transición entre ambas muestras. A esta
muestra combinada se le aplicaron los métodos estasticos
de vanguardia utilizados en la elaboración de los Patrones
de Crecimiento Infantil de la OMS (05 os), es decir, la
transformación de potencia de Box-Cox exponencial, junto con
instrumentos diagnósticos apropiados para seleccionar los mejores
modelos.
Resultados La fusión de los dos conjuntos de datos proporcionó
una transición suave de la talla para la edad, el peso para la edad
y el IMC para la edad a los 5 años. Con respecto al IMC para la
edad, la magnitud de la diferencia entre ambas curvas a los 5 años
fue generalmente de 0,0 kg/m² a 0,1 kg/m² en todos los centiles.
A los 19 años, los nuevos valores del IMC para +1 desviación
estándar (DE) fueron de 25,4 kg/m² para el sexo masculino y de
25,0 kg/m² para el sexo femenino, es decir, equivalentes al valor de
corte del sobrepeso en adultos (> 25,0 kg/m²). A su vez, el valor
correspondiente a +2 DE (29,7 kg/m² en ambos sexos) fue muy
similar al valor de corte de la obesidad (> 30,0 kg/m²).
Conclusión Las nuevas curvas se ajustan bien a los Patrones
de Crecimiento Infantil de la OMS a los 5 años y a los valores
de corte del sobrepeso y de la obesidad recomendados para los
adultos a los 19 años, colman la laguna existente en las curvas de
crecimiento y constituyen una referencia apropiada para el grupo
de 5 a 19 años de edad.



statistiques correspondant à l’état de la technique [méthode
Box-Cox-power-exponential (BCPE), complétée par des outils
permettant de sélectionner les meilleurs modèles], ayant servi à
construire la norme OMS de croissance de l’enfant (0 à 5 ans),
ont été appliquées à cet échantillon combiné.
Résultats La fusion des jeux de données a permis d’obtenir
une transition plus douce au niveau de 5 ans pour les courbes
de taille, de poids et d’IMC en fonction de l’âge. S’agissant de
l’IMC en fonction de l’âge, sur l’ensemble des centiles, l’ampleur
de la différence entre les deux courbes à l’âge de 5 ans se situe
principalement entre 0,0 kg/m² et 0,1 kg/m². A 19 ans, les
nouvelles valeurs d’IMC correspondant à un écart type de +1 sont
de 25,4 kg/m² pour les garçons et de 25,0 kg/pour les filles. Ces
valeurs concordent avec le point de coupure pour l’excès pondéral
chez l’adulte (> 25,0 kg/m²). De même, les valeurs correspondant
à plus de 2 écarts types (29,7 kg/m² pour les deux sexes) sont très
proches du point de coupure pour l’obésité (> 30,0 kg/m²).
Conclusion Les nouvelles courbes coïncident étroitement à 5 ans
avec la norme OMS de croissance de l’enfant et à 19 ans avec
les points de coupure recommandés chez l’adulte pour l’excès
pondéral et l’obésité. Elles comblent les lacunes en matière de
courbes de croissance et fournissent une référence appropriée
pour la tranche d’âges 5 -19 ans.






























        




667
Bulletin of the World Health Organization | September 2007, 85 (9)
Research
Growth curves for school-aged children and adolescents
Mercedes de Onis et al.
References
1. Lobstein T, Baur L, Uauy R. IASO International Obesity Task Force. Obesity in
children and young people: a crisis in public health.
Obes Rev
2004;5:4-104.
2. WHO Multicentre Growth Reference Study Group. WHO Child Growth
Standards based on length/height, weight and age.
Acta Paediatr Suppl
2006;450:76-85.
3. De Onis M. The use of anthropometry in the prevention of childhood
overweight and obesity.
Int J Obes Relat Metab Disord
2004;28:S81-5.
4. Physical status: the use and interpretation of anthropometry. Report of a
WHO Expert Committee.
World Health Organ Tech Rep Ser
1995;
854:161-262.
5. Wang Y, Moreno LA, Caballero B, Cole TJ. Limitations of the current World
Health Organization growth references for children and adolescents.
Food
Nutr Bull
2006;27:S175-88.
6. Must A, Dallal GE, Dietz WH. Reference data for obesity: 85th and 95th
percentiles of body mass index (wt/ht²) and triceps skinfold thickness.
Am J
Clin Nutr
1991;53:839-46.
7. Butte NF, Garza C, editors. Development of an international growth
standard for preadolescent and adolescent children.
Food Nutr Bull
2006;
27:S169-326.
8. Butte NF, Garza C, de Onis M. Evaluation of the feasibility of international
growth standards for school-aged children and adolescents.
J Nutr
2007;
137:153-57.
9. Borghi E, de Onis M, Garza C, Van den Broeck J, Frongillo EA, Grummer-
Strawn L, et al., for the WHO Multicentre Growth Reference Study Group.
Construction of the World Health Organization child growth standards:
selection of methods for attained growth curves.
Stat Med
2006;25:247-65.
10. 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:
WHO; 2006.
11. Hamill PV, Drizd TA, Johnson CL, Reed RB, Roche AF. NCHS growth curves for
children birth-18 years: United States.
Vital Health Stat 11
1977;165:i-iv,
1-74.
12. Rigby RA, Stasinopoulos DM. Smooth centile curves for skew and kurtotic
data modelled using the Box-Cox power exponential distribution.
Stat Med
2004;23:3053-76.
13. Cole TJ, Green PJ. Smoothing reference centile curves: the LMS method and
penalized likelihood.
Stat Med
1992;11:1305-19.
14. Dibley MJ, Goldsby JB, Staehling NW, Trowbridge FL. Development of
normalized curves for the international growth reference: historical and
technical considerations.
Am J Clin Nutr
1987;46:736-48.
15. Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Strawn LM, Flegal KM, Mei Z,
et al. 2000 CDC growth charts for the United States: methods and
development.
Vital Health Stat 11
2002;246:1-190.
16. Cole TJ, Freeman JV, Preece MA. British 1990 growth reference centiles for
weight, height, body mass index and head circumference fitted by maximum
penalized likelihood.
Stat Med
1998;17:407-29.
17. Fredriks AM, van Buuren S, Burgmeijer RJ, Meulmeester JF, Beuker RJ,
Brugman E, et al. Continuing positive secular growth change in the
Netherlands 1955-1997.
Pediatr Res
2000;47:316-23.
18. Obesity: preventing and managing the global epidemic. Report of a WHO
consultation.
World Health Organ Tech Rep Ser
2000;894:1-253.
19. Young people’s health — a challenge for society. Report of a WHO Study
Group on young people and Health for All by the Year 2000.
World Health
Organ Tech Rep Ser
1986;731:1-117.
... The weight and height of children were measured to calculate BMI using the formula BMI = weight (kg)/height 2 (m 2 ). Students were divided into 3 categories (wasting, normal, and overweight or obese) by using the BMI cutoff recommended by the World Health Organization [40]. For a detailed division, see Additional file 2: Supplement Table 2. ...
Article
Full-text available
Background Overweight and obesity rates have increased rapidly in Chinese school-age children, and previous studies have indicated that poor dietary literacy can lead to unhealthy eating behaviours. However, few studies have investigated the association between the dietary literacy of daily diet providers and the eating behaviours and nutritional status of school-age children raised by the providers. Thus, we aimed to explore this association. Methods We collected data on the eating behaviours and nutritional status of children in two primary schools in Anhui Province, as well as the dietary literacy of their daily diet providers. T-tests, one-way ANOVA, chi-square tests, and multiple linear regression were used to analyse the association. Results We found significant differences in the scores on the Questionnaire of Children's Daily Diet Providers' Dietary Literacy (QCDDPDL) by region, relationship with the child, age, and educational level of the daily diet provider (all p < .05). Moreover, the children in the low QCDDPDL score group were inclined to engage in unhealthy eating behaviours such as emotional undereating and overeating (p < .05). In addition, the incidence of overweight and obesity was higher in the low QCDDPDL attitude score group than in the high score group (p = .006). Conclusions Our study showed that the dietary literacy of diet providers may influence children's health and eating behaviours. Improving the dietary literacy of diet providers may promote the health status and eating behaviours of school-age children.
... Height and weight were measured while participants were dressed in light indoor clothing without footwear. BMI z-scores, adjusted for age and sex, were calculated (WHO references) [17]. ...
Article
Full-text available
Diet, screen time, physical activity, and sleep combine into lifestyle patterns with synergistic effects on health. This study aimed to identify lifestyle patterns in children without housing and assess their associations with physical and mental health and family socio-ecological factors. In the 2013 ENFAMS cross-sectional survey (children aged 6–12 experiencing homelessness, Greater Paris area, n = 235), parents reported socio-ecological factors, children’s behaviours, and mental health (the latter was also child-reported). Nurses measured children’s haemoglobin concentrations and body mass index. Principal component analysis was used to derive sex-specific lifestyle patterns. Hierarchical linear regressions and “outcome-wide” analyses assessed, respectively, these patterns’ relations to health and family socio-ecological factors. A rather healthy lifestyle pattern—similarly characterized by diverse diet and high sleep time—was identified, with slight differences by sex. Scores for this pattern were higher for children in food-secure or higher-income households, whose parents were proficient in French, who slept longer, or who received more social support compared to their counterparts, with some nuances by sex. Higher scores for this pattern were associated with higher prosocial behaviour scores (girls) and lower anxiety and hyperactivity–inattention symptoms scores (boys), but not with physical health. For this underserved and understudied population, the results highlight the importance of family socio-ecological factors in shaping the lifestyles and mental health of children.
... Height and weight Z-scores were calculated using the INMU-NutriStat software program (Mahidol University, 2002). Due to the lack of a national BMI reference, BMI Z-scores were analyzed using World Health Organization (WHO) 2007 growth reference data [16]. HbA1c was determined by turbidimetric inhibition immunoassay (Integra 400 analyzer; Roche Diagnostics). ...
Article
Full-text available
Appropriate dietary intake and physical activity (PA) are essential for glycemic control and optimal growth in youth with type 1 diabetes (T1D). Thus, this study aimed to compare dietary intake and PA between youth with T1D and healthy controls. One hundred Thai youth with T1D and 100 age-matched healthy participants were recruited. A 3-day food record was completed and converted into nutrient intake data. PA data were collected via interview. Participants with T1D had a significantly higher mean ± SD carbohydrate (50.8 ± 6.8% vs. 46.2 ± 7.5%, p < 0.01), lower fat (32.4 ± 5.9% vs. 35.9 ± 6.4%, p < 0.01), and lower protein (16.8 ± 2.6% vs. 17.9 ± 3.5%, p = 0.01) intake compared to controls. Fifty percent of T1D participants and 41% of control participants consumed saturated fat more than recommendations (p = 0.20). Participants with T1D had a higher median (IQR) calcium intake compared to controls (474 (297–700) vs. 328 (167–447) mg/day, p < 0.01). Both groups consumed less fiber and more sodium compared to recommendations. Both groups had inadequate PA. Participants with T1D had significantly less PA compared to controls (25 (13–48) vs. 34 (14–77) minutes/day, p = 0.04). In addition to the need for counseling that promotes consumption of more dietary fiber and calcium and less saturated fat and sodium, the benefits of performing regular exercise need to be emphasized among youth with T1D.
... Weight and height measurements were used to classify children according to three international standards: World Health Organization (WHO), CDC and the International Obesity Task Force (IOTF). The 2007 WHO Growth reference standards for children 2 years or older were used to estimate the body mass index (BMI) for sex and age z-score, which was then categorised as underweight (lower than −2 Standard Deviations [SD]), normal weight (between −2 SD and +1 SD), overweight (higher than +1 SD, which corresponds to a BMI of 25 kg/m 2 at 19 years), and obesity (higher than +2 SD, which corresponds to a BMI of 30 kg/m 2 at 19 years) [23,24]. Using the CDC's sex-specific 2000 BMI-for-age growth charts for the US child population, underweight was defined as a BMI-for-sex-and-age lower than the 5th percentile, normal weight as a BMI-for-sex-and-age equal or higher than the 5th percentile but lower than the 85th percentile, overweight as a BMI-for-sex-and-age equal or higher than the 85th percentile but lower than the 95th percentile, and obese as a BMI-for-sex-and-age higher than the 95th percentile [25]. ...
Article
Full-text available
Background: Childhood obesity and dental caries are prevalent chronic, multifactorial conditions with adverse health consequences and considerable healthcare costs. The aims of this study were: (1) to evaluate the relationship between obesity and dental caries among young children using multiple definitions for both conditions, and (2) to evaluate the role of family socioeconomic status (SES) and the child’s intake of added sugars in explaining this association. Methods: Data from 2775 2–5-year-olds children from the National Health and Nutrition Examination Survey (NHANES) 2011–2018 were analysed. Three different international standards were used to define obesity, namely the World Health Organization (WHO), Centers for Disease Control and Prevention (CDC), and the International Obesity Task Force (IOTF). Dental caries was measured during clinical examinations and summarised as counts (dt and dft scores) and prevalence (untreated caries [dt > 0] and caries experience [dft > 0]). The association of obesity with dental caries was assessed in regression models controlling for demographic factors, family SES and child’s intake of added sugars. Results: In crude models, obesity was associated with greater dt scores when using the IOTF standards (RR: 2.43, 95% CI: 1.11, 5.29) but not when using the WHO and CDC standards; obesity was associated with greater dft scores when using the WHO (1.57, 95%CI: 1.11–2.22), CDC (1.70, 95%CI: 1.17–2.46) and IOTF standards (2.43, 95%CI: 1.73–3.42); obesity was associated with lifetime caries prevalence when using the WHO (1.55, 95%CI: 1.05–2.29), CDC (1.73, 95%CI: 1.14–2.62) and IOTF standards (2.45, 95%CI: 1.61–3.71), but not with untreated caries prevalence. These associations were fully attenuated after controlling for demographic factors, family SES and child’s intake of added sugars. Conclusions: The relationship between obesity and dental caries in primary teeth varied based on the definition of obesity and dental caries used. Associations were observed when obesity was defined using the IOTF standards and dental caries was defined using lifetime indicators. Associations were fully attenuated after adjusting for well-known determinants of both conditions.
... /fpubh. . and World Health Organization (WHO) standard were also used to distinguish overweight and obesity (21,22). In our study, we used the above three standards to evaluate overweight and obesity prevalence in 2019, and used WGOC standard to assess the trends of overweight and obesity prevalence from 2000 to 2019. ...
Article
Full-text available
Objectives Overweight and obesity are harmful to human health. However, the latest trends of Chinese childhood overweight and obesity prevalence are not available. The aim of this study was to examine the trends from 2000 to 2019 among students in China. Methods We analyzed data of 66,072 students in the Chinese National Survey on Students' Constitution and Health from 2000 to 2019. Overweight and obesity were defined based on the standard formulated by the International Obesity Task Force (IOTF standard), the World Health Organization (WHO standard), and the Working Group on Obesity in China (WGOC standard), respectively. The χ ² -test was used to test the trends of overweight and obesity prevalence and logistic regression was conducted to evaluate the prevalence odds ratios of boys vs. girls and urban vs. rural areas. Results The prevalence of obesity/overweight and obesity combined was 6.03/23.58% (IOTF standard), 10.56/25.88% (WGOC standard) and 10.75/29.69% (WHO standard) in 2019. From 2000 to 2019, according to the WGOC standard, the prevalence increased from 2.51 to 10.56% for obesity and increased from 9.81 to 25.88% for overweight and obesity combined ( P for trend < 0.001). Obesity/overweight and obesity were greater problems in boys than girls and urban than rural areas, but urban-rural differences decreased over time. Conclusion Overweight and obesity prevalence increased significantly in children and adolescents in China from 2000 to 2019. The prevalence of overweight and obesity in rural areas may contribute to a large percentage of children with overweight and obesity.
... We also measured umbilical waist circumference (WC, cm) as an indicator of abdominal adiposity [21,22]. Z-scores for weight, height, and BMI were calculated using the 2007 World Health Organization (WHO) values as the reference [23], classifying the children as thin (<−2 SD), normal weight (between −2 SD and 1 SD), overweight (between 1 and 2 SD) and obese (≥2 SD). ...
Article
Full-text available
Background: The aim of this study was to evaluate the relationship between the Dietary Inflammatory Index (DII®) and cardiovascular health indicators in children. Methods: The sample consisted of 365 schoolchildren aged 8 to 12 from the Region of Madrid. Anthropometric and hemodynamic measurements were collected. Variables relating to habits and lifestyles, parental level of education, and data on their diet, through three 24 h food recall surveys, were also collected. The diet quality indicators considered are the DII based on 25 nutrients and the KIDMED index. Results: Children with a more pro-inflammatory diet came from families with lower levels of parental education (p < 0.05). Predictive models show that in the group with a more pro-inflammatory diet (>P50), the likelihood of developing hypertension in childhood is 2.1 times higher (OR = 2.085 (1.107–3.927)) and they have more than twice the risk of developing obesity (OR = 2.3) or developing obesity and hypertension simultaneously (OR = 1.290 (1.316–3.985)). Furthermore, predictive models showed that the children with a pro-inflammatory diet (>P50) had higher values for BFM% (β = 1.957; p = 0.026) and BMI (β = 0.015; p = 0.012) than children with a lower inflammatory diet (<P50). Conclusions: Higher values on the DII are related to poorer nutritional status and cardiovascular health in childhood. Thus, a pro-inflammatory diet is also associated with a lower socio-economic level and poorer diet quality.
... Índice de masa corporal (IMC): Se les preguntó la altura y el peso, se calculó el IMC (Kg/m2), y se clasificó a la muestra según las tablas de IMC para niños y niñas de 5 a 18 años de edad y adultos no embarazadas, no lactantes (de Onis et al., 2007). ...
Chapter
Full-text available
Exploramos las variables psicosociales relacionadas con el control de peso en adolescentes de México y España. Estos estudios ponen de manifiesto que la motivación controlado, bajo sentido de eficacia y miedo a engordar son factores que llevan a estos adolescentes a iniciar el control de peso. We explored psychosocial variables related to weight control in adolescents from Mexico and Spain. These studies show that controlled motivation, low sense of efficacy, and fear of gaining weight are factors that lead these adolescents to initiate weight control.
Article
Introduction: Cardiometabolic risk factors often emerge in childhood, increasing the risk of cardiovascular diseases in adulthood. The school environments represent an important active space in a child's routine and may influence their health status. Methods: In this cross-sectional study, we aimed to evaluate the clustering of cardiometabolic risk factors and its association with the obesogenic and leptogenic environment around schools. A total of 378 children (181 boys; 197 girls) aged 8 and 9 years, enrolled in all urban schools of Viçosa, Minas Gerais, Brazil, were evaluated. Concentrations of high-density lipoprotein-cholesterol (HDL-c) and triglycerides (TG), and the insulin resistance (HOMA-IR), waist circumference (WC), and mean arterial pressure (MAP) were measured. The neighborhood income, walkability index, predominantly ultra-processed food stores, public spaces for leisure and/or physical activities, traffic accidents, crime and green spaces densities were assessed in 400 road network buffers around schools. To test association of the obesogenic and leptogenic environment around schools with cardiometabolic risk clustering, binary logistic regression models were performed with generalized estimating equations. Results: "Obesogenic school environments" had greater densities of ultra-processed food stores, crime and traffic accidents, and higher walkability. The "↓ Atherogenic risk" cluster consisted of lower TG and higher TG values. There was an inverse association between the obesogenic environment around schools and a child's "↓ atherogenic risk" clustering (OR = 0.63; p < 0.001). Conclusion: We concluded that public policies aimed at modifying the environment around schools, by providing healthier food options and safe conditions for active mobility, are essential to prevent child's atherogenic risk.
Article
Full-text available
How well students learn and perform in academic contexts is a focus of interest for the students, their families, and the entire educational system. Although evidence has shown that several neurobiological factors are involved in scholastic achievement (SA), specific brain measures associated with academic outcomes and whether such associations are independent of other factors remain unclear. This study attempts to identify the relationship between brain structural parameters, and the Chilean national University Selection Test (PSU) results in high school graduates within a multidimensional approach that considers socio-economic, intellectual, nutritional, and demographic variables. To this end, the brain morphology of a sample of 102 students who took the PSU test was estimated using Magnetic Resonance Imaging. Anthropometric parameters, intellectual ability (IA), and socioeconomic status (SES) were also measured. The results revealed that, independently of sex, IA, gray matter volume, right inferior frontal gyrus thickness, and SES were significantly associated with SA. These findings highlight the role of nutrition, health, and socioeconomic variables in academic success.
Article
Several studies have reported that the atherogenic index of plasma ( AIP) is associated with serum uric acid (SUA) in healthy middle-aged men, kidney transplant patients, and diabetic patients. We assessed the relationship between AIP and SUA in patients with untreated essential hypertension (EH) in China. In the cross-sectional study, we evaluated 5003 patients with untreated EH in China; 1490 were assigned to the hyperuricemia (HUA) group and 3513 to the normouricemia (non-HUA) group according to the diagnostic criteria of HUA (SUA level≥6 mg/dL in females and ≥7 mg/dL in males). A higher AIP level was found in the HUA group ( P < .001), and AIP was positively and independently correlated with SUA after adjusting for statistically significant variables (β = .204, P < .001). Moreover, multivariate logistic regression analysis showed that the AIP was closely and independently related to HUA after adjusting for confounders (odds ratio [OR]: 11.980, 95% confidence interval [CI]: 3.186–45.047, P < .001). Our findings suggested that the AIP was positively and independently associated with SUA, and elevated AIP might be a significant risk factor for HUA in patients with untreated EH in China.
Article
Full-text available
Body mass index (BMI) and triceps skinfold thickness (TSF) are commonly used measures of adiposity in clinical and epidemiologic studies. The 85th and 95th percentiles of BMI and TSF are often used operationally to define obesity and superobesity, respectively. Race-specific and population-based 85th and 95th percentiles of BMI and TSF for people aged 6-74 y were generated from anthropometric data gathered in the National Health and Nutrition Examination Survey 1 (NHANES I). The complex sample design of the survey is reflected in the reference values presented. Racial differences in these extremes of the distribution do not emerge until adulthood. Researchers may choose population-based, race-specific, or age-specific criteria for obesity on the basis of assumptions underlying their specific research questions.
Article
Full-text available
The World Health Organization recommended in 1978 that the National Center for Health Statistics/Centers for Disease Control growth reference curves be used as an international growth reference. To permit the expression of growth in terms of standard deviations, CDC developed growth curves from the observed data that approximate normal distributions. Because of significant skewness, standard deviations for weight-for-age and weight-for-height were calculated separately for distributions below and above the median. Standard deviations below the median were calculated from the 5th, 10th, 25th, and 50th observed percentiles while those above the median were based on the 50th, 75th, 90th, and 95th observed percentiles. Height-for-age distributions did not show significant skewness, thus, the standard deviations were calculated based on all six of the above observed percentiles. The normalized reference curves provide a highly useful data base that permits the standardized comparison of anthropometric data from different populations.
Article
Aim: To describe the methods used to construct the WHO Child Growth Standards based on length/height, weight and age, and to present resulting growth charts. Methods: The WHO Child Growth Standards were derived from an international sample of healthy breastfed infants and young children raised in environments that do not constrain growth. Rigorous methods of data collection and standardized procedures across study sites yielded very high-quality data. The generation of the standards followed methodical, state-of-the-art statistical methodologies. The Box-Cox power exponential (BCPE) method, with curve smoothing by cubic splines, was used to construct the curves. The BCPE accommodates various kinds of distributions, from normal to skewed or kurtotic, as necessary. A set of diagnostic tools was used to detect possible biases in estimated percentiles or z-score curves. Results: There was wide variability in the degrees of freedom required for the cubic splines to achieve the best model. Except for length/height-for-age, which followed a normal distribution, all other standards needed to model skewness but not kurtosis. Length-for-age and height-for-age standards were constructed by fitting a unique model that reflected the 0.7-cm average difference between these two measurements. The concordance between smoothed percentile curves and empirical percentiles was excellent and free of bias. Percentiles and z-score curves for boys and girls aged 0-60 mo were generated for weight-for-age, length/height-for-age, weight-for-length/h eight (45 to 110 cm and 65 to 120 cm, respectively) and body mass index-for-age. Conclusion: The WHO Child Growth Standards depict normal growth under optimal environmental conditions and can be used to assess children everywhere, regardless of ethnicity, socio-economic status and type of feeding.
Article
Since 1858, an increase of mean stature has been observed in the Netherlands, reflecting the improving nutritional, hygienic, and health status of the population. In this study, stature, weight, and pubertal development of Dutch youth, derived from four consecutive nationwide cross-sectional growth studies during the past 42 y, are compared to assess the size and rate of the secular growth change. Data on length, height, weight, head circumfer- ence, sexual maturation, and demographics of 14 500 boys and girls of Dutch origin in the age range 0 -20 y were collected in 1996 and 1997. Growth references for height and weight were constructed with a method that summarizes the distribution by three smooth curves representing skewness (L curve), the median (M curve), and coefficient of variation (S curve). The relationship between height and demographic variables was assessed by multivariate analysis. Reference curves for menarche and sec- ondary sex characteristics were estimated by a generalized addi- tive model using a logit transformation. A positive secular growth change has been present in the past 42 y for children, adolescents, and young adults of Dutch origin, although at a slower rate in the last 17 y. Height differences according to region, educational level of child and parents, and family size have remained. In girls, median age at menarche has decreased by 6 mo during the past four decades to 13.15 y. Environmental conditions have been favorable for many decades in the Nether- lands, and the positive secular change in height has not yet come to a halt, in contrast to Scandinavian countries. Main contributors to the increase in height may be improved nutrition, child health, and hygiene, and a reduction of family size. (Pediatr Res 47: 316-323, 2000)
Article
Refence centile curves show the distribution of a measurement as it changes according to some covariate, often age. The LMS method summarizes the changing distribution by three curves representing the median, coefficient of variation and skewness, the latter expressed as a Box-Cox power. Using penalized likelihood the three curves can be fitted as cubic splines by non-linear regression, and the extent of smoothing required can be expressed in terms of smoothing parameters or equivalent degrees of freedom. The method is illustrated with data on triceps skinfold in Gambian girls and women, and body weight in U.S.A. girls.