Trends in height and BMI of 6-year-old children during the nutrition transition in Chile.

Abstract

We analyzed trends in height and BMI and their interaction in 6-year-old Chilean children over the last 15 years.
We calculated height for age z-score (HAZ), BMI z-score, prevalence of obesity, underweight, and stunting from cross-sectional national school-based annual population surveys in 1987, 1990, 1993, 1996, 2000, and 2002. Using mixed model analysis, we determined the risk of obesity according to height over time as odds ratios (ORs) and 95% confidence interval and the potential influence of height and year of study on BMI z-score.
Over the study period, height increased by 2.8 cm in boys and 2.6 cm in girls, whereas stunting declined from 5% to 2% in both. Tallness increased by approximately 2%, BMI z-score increased from +0.3 to +0.65 in boys and to +0.62 in girls, and HAZ increased from -0.47 in boys and -0.45 in girls to 0 in 2002. Underweight declined from 4% to 3%, whereas obesity rose from 5% to approximately 14%. The probability of obesity among tall children was significantly greater than that for normal height children (OR, 2.3 to 3.5). The lowest obesity risk was observed between -2 and -1 HAZ. The OR for obesity in the stunted relative to normal height children was variable, ranging from 1.23 to 0.65, whereas it was significant and consistently positive (1.1 to 1.7) for boys and girls when it was compared with the lowest obesity risk according to height.
Tallness is significantly associated with increased obesity risk in children, while stunting is also associated, but to a lesser degree.

Full text

Page 1

ORIGINAL COMMUNICATION
Trends in overweight and obesity prevalence in
Chilean children: comparison of three definitions
J Kain1*, R Uauy1, F Vio1and C Albala1
1Institute of Nutrition and Food Technology (INTA), University of Chile, Santiago, Chile
Objective: To compare trends in prevalence of overweight and obesity among Chilean children using three different criteria.
Methods: Descriptive cross-sectional school-based study to analyse these trends in 6-y-old boys and girls who entered first grade
in 1987, 1990, 1993, 1996 and 2000. Gender-specific prevalence of overweight and obesity were determined with three
criteria: weight for height (W-H) Z-scores compared to NCHS 1977; present reference used by the Ministry of Health; and body
mass index (BMI) compared to the revised US CDC Growth Charts with cut-off values of P85-95 and P-95 and IOTF reference
with cut-offs extrapolated from an adult BMI of 25 and 30.
Results: The prevalence of overweight determined by W-H (WHO) increased from 15% in 1987 to 20% in 2000 for boys and
from 17.2 to 21.8% for girls. With BMI-CDC, the increase was from 13.2 to 19.2% for boys and 12 to 18.5% for girls. With BMI-
IOTF, rates were very similar. Prevalence of obesity using W-H (WHO) increased from 6.5% in 1987 to 17% in 2000 for boys and
from 7.8 to 18.6% for girls. Using BMI-CDC, the increase was from 5.1 to 14.7% for boys and from 4 to 15.8% for girls; using
BMI-IOTF prevalence estimates were much lower.
Conclusions: Obesity prevalence in children has increased over time, and trends are similar independent of criteria. The
reference used to define prevalence is important since it provides different estimates. Policy makers should be aware that a
spurious drop in prevalence may appear if the IOTF reference is compared to the other criteria.
European Journal of Clinical Nutrition (2002) 56, 200–204. DOI: 10.1038=sj=ejcn=1601301
Keywords: children; obesity; trend; BMI; weight for height; prevalence
Introduction
Obesity has been recognized as a major public health pro-
blem; its prevalence has increased both in industrialized and
in developing countries (World Health Organization, 1998).
In Chile, it has markedly increased both in adults and in
children. Data from a representative sample of the adult
population living in Santiago shows that, in 1992, 11% of
men and 24% of women were classified as obese, defined by a
body mass index (BMI) ?30, compared to 6 and 14%,
respectively, in 1988 (Vio & Albala, 2000). In children,
there has been an important decline in the prevalence of
undernutrition, concurrently with spiraling obesity rates
(Kain et al, 1998). The reasons behind the important reduc-
tion in numbers of children with nutritional deficit are most
likely found in existing social programmes, including food
intervention. Greater access to an improved educational
system, progress in water and sanitation infrastructure,
wide coverage of primary health care interventions and
declining unemployment rates are also major contributors
(Uauy et al, 2001).
The improvement in the economic situation and the
‘modernization’ of society has led to an increase in the
consumption of high-energy foods and an alarming increase
in sedentary behaviour. These are probably the main reasons
why obesity rates in pre-school and schoolchildren more
than doubled over the past decade (Kain & Andrade, 1999).
The terms overweight and obesity are often used inter-
changeably, despite the fact that they are not identical.
Overweight is defined as an increased weight (not necessarily
excess fat) for a certain height, while obesity indicates an
excess in fat mass (Troiano & Flegal, 1998; Barlow & Dietz,
1998). Even though the long-term effect of overweight and
obesity on morbidity and mortality in children has not yet
*Correspondence: J Kain, Institute of Nutrition and Food Technology
(INTA), University of Chile, Casilla 138–11. Santiago, Chile.
E-mail: jkain@uec.inta.uchile.cl
Contributors: J Kain, R Uauy, F Vio and C Albala.
Received 26 March 2001; revised 29 June 2001;
accepted 3 July 2001
European Journal of Clinical Nutrition (2002) 56, 200–204
? 2002 Nature Publishing Group All rights reserved 0954–3007/02 $25.00
www.nature.com/ejcn

Page 2

been as well documented as in adults, multiple studies have
shown that adiposity in childhood is correlated with the
rising incidence of diabetes, hypertension and atheroesclero-
sis observed in this age group (Power et al, 1997; Freedman
et al, 1999; Berenson et al, 1998; Dietz, 1998). The conse-
quences of overweight and obesity imply that the definition
of who is overweight, and especially who is obese, is of
utmost importance.
In 1992 the Chilean Ministry of Health adopted the World
Health Organization’s international reference (Bulletin of
the WHO, 1986) and the weight for height index (W–H)
as the official criterion to evaluate the nutritional status of
pre-school children both in terms of undernutrition and
overnutrition. The Chilean Ministry of Education, which
collects data on weight and height of schoolchildren in the
first and ninth grades, also adopted this criteria for children
under 10y of age.
Presently the scientific community has recommended the
use of the BMI to classify overweight and obese children
from the age of 2 (Dietz & Bellizzi, 1999; Pietrobelli et al,
1998; Dietz & Robinson, 1998; Gallagher et al, 1996).
Depending on the criteria used, obesity prevalence gives
very different results.
The purpose of this paper is to compare prevalence esti-
mates of overweight and obesity among Chilean school-
children from 1987 until 2000 determined by three
different criteria.
Methods
Sample
This is a descriptive population-based study whose purpose
was to analyse cross-sectionally the prevalence of overweight
and obesity among schoolchildren who entered first grade in
1987, 1990, 1993, 1996 and 2000. We analysed data from
children entering primary education, because the school
census collects anthropometric data only for children in
the first and ninth grades. Data for ninth graders are not
available before 1993. Weight and height of children in the
first grade were collected by the teacher of each school in
March, the beginning of the school year. Teachers were,
instructed on how to fill in general information on the
children and how to weigh and measure them. Additionally,
they learn how to calibrate the scale, which was a standard
physician’s beam scale. Children were measured with light
clothes and without shoes. Weight was recorded to the
nearest 0.25kg and height using a stadiometer to the nearest
0.5cm following a standardized protocol. True precision is
likely to be less due to measurement error. The socio-eco-
nomic situation of the country over the study period
improved substantially; average yearly economic growth
was around 6–7%. In 1987, the proportion of the population
classified as poor (based on the purchasing power of a basic
food basket) was 44%, in 1998 it dropped to 23% (Chilean
Ministry of Planning, 1999). The population remained pre-
dominantly urban over the period, around 85%. Children in
first grade from state-supported schools constitute over 91%
of all children; the rest are in private schools.
The anthropometric census provided data on weight and
height for all the children entering primary education in
state-supported schools. We included in our analysis, those
aged 71–83 months; these comprised approximately 83% of
the total in first grade from those schools. Other ages were
excluded since children who enter school at older ages
present subnormal heights, thus the effect of stunting may
confound the interpretation of the data (Kain et al, 1999).
The absolute number of children considered in the analyses,
according to gender and year of study is shown in Table 1.
Analysis
The prevalence of overweight and obesity was calculated
using two indicators, W-H and BMI (kg=m2). The W-H
index (compared to the WHO reference) was adopted by
the Ministry of Health to determine the nutritional status of
pre-school children (2–5y of age) who were beneficiaries of
the Primary Health Care System. The Ministry of Education
also uses this definition for its anthropometric census on
schoolchildren in first grade. Overweight was defined as a W-
H between > þ1 and þ2s.d., while obesity was defined as a
W-H> þ2s.d. of the reference (Bulletin of the WHO, 1986).
We also determined overweight and obesity using BMI,
because of the endorsement given by the international
scientific community to the use of this index to classify
both these conditions in children at the population level.
The BMI data were analysed using two different reference for
comparison, the revised CDC Growth Charts (CDC=NCHS,
2000) and the international reference provided by the IOTF
(Cole et al, 2000). Although the CDC growth charts were
developed to evaluate the nutritional status of US children,
we considered it useful to compare our official prevalence
estimates with the latest existing reference based on large
representative datasets. The revised CDC reference originates
from five cross-sectional representative surveys carried out in
the US between 1963 and 1994; the most prominent feature
is the addition of the BMI-for-age curves. In this respect, it is
important to point out that these curves exclude the data for
children ? 6y of age from the last survey (NHANES III). The
cut-off values considered for overweight and obesity fol-
lowed the advice of the US Expert Committee on Childhood
Obesity (Barlow & Dietz, 1998), which recommended that
for children and adolescents overweight be defined as a BMI
between the 85th and 95th percentile for age and sex, and
obesity, a BMI ? 95th percentile. However, BMI values
Table 1
analysis
Number of schoolchildren in first grade considered in the
19871990199319962000
Boys
Girls
85259
81632
92685
89075
100353
96759
101482
96356
102316
97128
Overweight and obesity in Chilean children
J Kain et al
201
European Journal of Clinical Nutrition

Page 3

>85th percentile are sometimes considered to indicate over-
weight (Flegal et al, 2001). The range of BMI cut-off values for
overweight and obesity for ages 71–83 months is shown in
Table 2. Distributions of month-specific BMIs were used to
determine overweight and obesity prevalence by gender
(CDC=NCHS, 2000).
The IOTF definition is based on the recommendations of
the Childhood Obesity Working Group convened in 1997,
which proposed that the adult BMI cut-off points of 25 and
30 (values adopted by WHO; Troiano & Flegal, 1998) be
linked to BMI centiles for children in order to provide
child cut-offs points. These two levels correspond to an
adult BMI of 25 and 30 and reflect values in children tracking
to overweight and obesity in adults (Cole et al, 2000; Flegal
et al, 2001). This reference, which also uses age—sex-specific
BMI centiles, is based on six large international cross-sec-
tional representative datasets, identifying the BMI values
that extrapolate to childhood. Reference values are given
in 6 month intervals. The range of BMI cut-offs using the
IOTF definition of overweight and obesity is given in Table 2.
Results
Tables 3 and 4 show the results of the overweight and obesity
prevalence for 6-y-old boys and girls using the three criteria
described under Methods and the percentage increment over
the 13y period. The overweight prevalence in boys deter-
mined by W-H (Table 3), increased from 15% in 1987 to 20%
in 2000; when derived using BMI, it increased from 13.2 to
19.2% based on NCHS=CDC criteria and from 10.6% to
18.8% if the IOTF criteria was used. This prevalence in girls
(Table 4), determined by W-H, increased from 17.2 to 21.8%.
When derived using BMI, it increased from 12 to 18.5%
based on NCHS=CDC, and from 11.6 to 19.6% when com-
pared with IOTF. Over the 13y period the rate of increment
in boys fluctuated between 33 and 77%. In girls, the differ-
ences followed similar trends.
The prevalence of obesity in boys (Table 3) increased from
6.5% in 1987 to 17% in 2000 when determined by W-H
(WHO); these figures were slightly lower if derived using
BMI-CDC, while with BMI-IOTF these were much lower—
1.8% in 1987 and 7.2% in 2000. Over the 13y period, the
Table 2
BMI cut-offs for schoolchildren in first grade
IOTF CDC=NCHS
Gender
Age range
(in months)
BMI range
for overweight
BMI range
for obesity
BMI range
for overweight
BMI range
for obesity
Male
Female
71–83
71–83
17.55–17.82
17.34–17.75
19.78–20.63
19.65–20.51
16.97–17.32
17.03–17.52
18.31–19.01
18.72–19.52
Table 4
Prevalence of overweight and obesity among 6-y-old Chilean girls defined by three criteria (1987–2000)
1987
%
1990
%
1993
%
1996
%
2000
%
Percentage increment
over 13y
Index and reference to define overweight
W-H>1 and ? s.d. WHO
BMI-CDC P85–P95
BMI-IOTF corresponding to an adult BMI 25–29.9
17.2
12.0
11.6
19.7
14.6
14.3
20.7
16.4
16.9
21.4
17.8
18.8
21.8
18.5
19.6
27
54
69
Index and reference to define obesity
W-H>2s.d. WHO
BMI-CDC ?P95
BMI-IOTF corresponding to an adult BMI ? 30
7.8
4.0
2.1
10.1
5.3
2.8
12.7
8.1
4.6
15.0
10.0
5.9
18.6
15.8
7.5
138
295
257
Table 3
Prevalence of overweight and obesity among 6-y-old Chilean boys defined by three criteria (1987–2000)
1987
%
1990
%
1993
%
1996
%
2000
%
Percentage increment
over 13y
Index and reference to define overweight
W-H>and ?2s.d. WHO
BMI-CDC P85–P95
BMI-IOTF corresponding to an adult BMI 25–29.9
15.0
13.2
10.6
17.3
15.4
13.4
18.2
16.9
15.9
19.0
18.2
17.8
20.0
19.2
18.8
33
45
77
Index and reference to define obesity
W-H>5s.d. WHO
BMI-CDC ?P95
IOTF-BMI corresponding to an adult BMI ?30
6.5
5.1
1.8
8.9
7.0
2.5
11.4
10.1
4.3
13.4
11.8
5.2
17.0
14.7
7.2
161
189
300
Overweight and obesity in Chilean children
J Kain et al
202
European Journal of Clinical Nutrition

Page 4

percentage increment was extremely high, namely, 161, 189
and 300%, respectively. Obesity in girls (Table 4) increased
from 7.8 to 18.6% with the first criterion, from 4 to 15.8%
with BMI-CDC, and from 2.1 to 7.5% with BMI-IOTF. These
increments follow similar trends as observed for boys.
The analysis by gender shows that the overweight pre-
valence is slightly higher in girls, while obesity is similar
when determined with BMI-IOTF, while it is consistently
higher among girls only when determined with W-H.
Discussion
The results of this comparative analysis show that the
absolute prevalence of overweight and obesity vary accord-
ing to the criteria used, and although trends increase mark-
edly over time, these rates are different. Although this is
obvious, it is disconcerting for program planners and the
general public. Presently, no reference to measure childhood
obesity has been validated prospectively in terms of predict-
ing health consequences for given BMI cut-off points. We
used the traditional international WHO reference and two
very recent BMI references proposed by reputable groups.
Troiano and Flegal, (1999) also described how different
definitions yielded divergent estimates from surveys carried
out in the US. For example, they report that the prevalence
of overweight among youth analysed from the third
National Health and Nutrition Examination Survey, ranged
from 11 to 24% depending on the criteria used. The main
difference arose from using the 85th percentile or the 95th
percentile as cut-off points for a given reference, rather than
from the use of a specific reference. Flegal et al (2001) in a
recent analysis comparing the prevalence of overweight and
obesity (referred to as levels 1 and 2) of US children 2–19y
of age included in the NHANES III (1988–1994) using
CDC=NCHS and IOTF, obtained similar rates with these
two references for overweight, but lower rates with IOTF
for the 6 to 8-y-olds. For boys in that age group, the
prevalence for overweight was 23.3% with CDC=NCHS and
18.3% with IOTF, while for obesity, these figures were 10.8
and 7.8% respectively. The rates for girls were similar. Our
results on obesity prevalence for 6-y-olds show similar rates
with CDC=NCHS in 1996, but lower rates with IOTF (around
5%). Flegal et al (2001) found that the differences between
methods were not systematic.
The WHO and the new NCHS=CDC references are based
on the distribution of representative samples of the US
population and use statistically based cut-off points, when
different measures are used to determine the prevalence of
overweight and obesity. In contrast, the IOTF reference
considers BMI cut-off points that are unrelated to the true
distribution of the reference populations. In fact, child cut-
off points are those which are extrapolated from a BMI of 25
and 30 at age 18; it is assumed that children with those BMI
present inherent health risk. The difference in the origin of
those cut-off points may explain in part why the absolute
prevalence of overweight and obesity are different.
Reilly et al (2000) in a recent analysis aimed at determin-
ing the sensitivity and specificity of both the BMI>95th
percentile (UK 1990 reference) and the BMI cut-off provided
by IOTF on 7-y-old children, concluded that the definition of
obesity using the IOTF criteria had low sensitivity. As
pointed out by the authors, obesity prevalence in epidemio-
logical studies will be substantially underestimated when
using BMI-IOTF. Our results confirm their suggestion, for
example in 6-y-olds the absolute prevalence of obesity is
reduced by half when using BMI-IOTF.
In general, obesity was slightly higher among girls; this is
mostly the case in developing countries as well as in devel-
oped ones (World Health Organization, 1998). Martorell et al
(2000) in a recent analyses of 71 datasets from developing
countries, found that the proportion of overweight was
significantly higher among girls in 11 countries and slightly
higher in the rest.
The influence of stunting in raising the prevalence of
obesity, given the effect of short stature on BMI needs to
be considered (Schroeder & Martorell, 1999; Trowbridge et al,
1987). We have observed a drop in stunting rates in Chile
concomitant to the rise in obesity. Thus the contribution of
stunting to the rise in obesity prevalence would have
decreased over time, suggesting that other factors are respon-
sible for the obesity trends. The prevalence of stunting
(height for age < 72Z WHO) in 6-y-olds was 10.6 and
7.3% for boys and girls, respectively, in 1987, while in
1996 these figures declined to 5.8 and 3.4%. In fact, the
average stature in this age group increased by 2cm in 10y
(Kain et al, 1999).
Weight for height to define obesity with cut-off points
based on standard deviations (or Z scores) of the WHO
reference (Gorstein et al, 1994; Habicht & Pelletier, 1990;
Dibley et al, 1987) is used to determine prevalence estimates
in Chile, thus this criterion was included in our evaluation.
We also used BMI, because it better reflects excess body fat in
children over 5y of age and most important, it explicitly
considers age (Poskitt, 1995; Ellis et al, 1999). Also, BMI-for-
age presents the smallest height bias in children. Addition-
ally, new information indicates that BMI is more frequently
used around the world (Gillaume, 1999; Bellizzi & Dietz,
1999). Despite the general acceptance of BMI, it is important
to note that the correlation coefficient of BMI with adiposity
is highly variable; it ranges from 0.4 to 0.9 according to the
age and gender of children (Dietz & Robinson, 1998; Poskitt,
1995). Also, BMI has limitations as a means of monitoring
childhood obesity because, being a proxy for both fat-free
and fat mass, it is not readily distinguishable to what extent
each of these compartments is affected by an increase in BMI
(Reilly et al, 2000).
With respect to the references, the fact that the new
NCHS=CDC BMI-for-age chart excludes data on weight of
children over 6y of age from the latest survey (NHANES III)
to avoid the influence of the increasing obesity prevalence
that has occurred among US children, may support its
applicability in countries which have not yet faced the
Overweight and obesity in Chilean children
J Kain et al
203
European Journal of Clinical Nutrition

Page 5

fully blown obesity epidemic. On the other hand, Cole et al,
2000) argue that the reference they published is less arbitrary
and more international than others and recommend its use
in international comparisons. In the near future, programme
planners in charge of nutrition programmes in Chile and
elsewhere most likely will adopt BMI to evaluate obesity in
pre-pubertal children. The choice of a reference and a cut-off
point will determine the absolute prevalence of obesity and
its trend. Unfortunately, there is no gold standard to define
obesity in terms of the degree of adiposity in children
affecting health outcomes throughout the lifecycle. Local
age-specific BMI references and cut-offs for children and
adolescents which provide predictive risks of significant
metabolic morbidity and disease burden throughout the
lifecycle are needed in order to define obesity and its con-
sequences across cultures.
Acknowledgements
The authors want to thank Drs Tim Cole and Ed Frongillo
who reviewed an earlier version of this manuscript.
References
Barlow S & Dietz W (1998): Obesity evaluation and treatment: expert
committee recommendations. Pediatrics 102, E29–E36.
Bellizzi M & Dietz W (1999): Workshop on childhood obesity:
summary of the discussion. Am. J. Clin. Nutr. 70, 173S–175S.
Berenson G, Srinivasan S & Nicklas T (1998): Autherosclerosis: a
nutritional disease of childhood. Am. J. Cardiol. 82, 22T–29T.
Bulletin of the WHO (1986): Use and interpretation of anthropometric
indicators of nutritional status. Geneva: WHO.
CDC=NCHS (2000)
CDC growth
www.cdc.gov/growthcharts. Posted May 30, 2000 on the Internet.
Chilean Ministry of Planning (1999): Poverty and income distribution in
Chile 1998. Santiago: Results from the National Socioeconomic
Characterization Survey.
Cole T, Bellizzi M, Flegal K & Dietz W (2000): Establishing standard
definition for child overweight and obesity worldwide: interna-
tional survey. Br. Med. J. 320, 1–6.
Dibley M, Staehling N, Nieburg P, Trowbridge F et al (1987): Intepre-
tation of Z-scores anthropometric indicators derived from the
international growth reference. Am. J. Clin. Nutr. 46, 736–748.
Dietz W (1998): Health consequences of obesity in youth: childhood
predictors of adult disease. Pediatrics 101(Suppl), 518–525.
Dietz W & Bellizzi M (1999): The use of body index to assess obesity
in children. Am. J. Clin. Nutr. 70(Suppl): 123S–125S.
Dietz W & Robinson T (1998): Use of body mass index (BMI) as a
measure of overweight in children and adolescents. J. Pediatr. 132,
191–193.
Ellis K, Abrams S & Wong W (1999): Monitoring childhood obesity:
assessment of the weight=height2index. Am. J. Epidemiol. 159,
939–946.
Flegal K, Ogden C, Wei R, Kuczmarski R & Johnson C (2001):
Prevalence of overweight in US children: comparison of US
growth charts from CDC with other reference values for body
mass index. Am. J. Clin. Nutr. 73, 1086–1093.
Freedman D, Dietz W, Srinivasan S & Berenson G (1999): The
relationship of overweight to cardiovascular risk factors among
children and adolescents: the Bogalusa Heart Study. Pediatrics 103,
1175–1182.
charts:United States.
http://
Gallagher D, Visser M, Sepulveda D, Pierson R, Harris T & Heymsfield
S (1996): How useful is body mass index for comparison of body
fatness across age, sex, and ethnic groups? Am. J. Epidemiol. 143,
228–239.
Gorstein J, Sullivan K, Yip R, de Onis M et al (1994): Issues in the
assessment of nutritional status when using anthropometry. Bull
WHO 72, 273–284.
Guillaume M (1999): Defining obesity in childhood: current practice.
Am J. Clin. Nutr. 70(Suppl): 126S–130S.
Habicht J-P & Pelletier D (1990): The importance of context in
choosing nutritional indicators. J. Nutr. 120, 1519–1524.
Kain J & Andrade M (1999): Characteristics of the diet and patterns of
physical activity in obese Chilean preschoolers. Nutr. Res. 19,
203–215.
Kain J, Vio F & Albala C (1998): Childhood nutrition in Chile: from
deficit to excess. Nutr. Res. 18, 1825–1835.
Kain J, Uauy R, Dı ´az M & Aburto A (1999): Height increase in
schoolchildren entering first grade during the last decade. Rev.
Med. Chil. 127, 539–546.
Martorell R, Kettel-Khan L, Hughes M & Grummer-Strawn L (2000):
Obesity in children from developing countries: a public health
problem? Presented at the 47th Nestle Workshop ‘Nutrition and
Growth’, Santiago, Chile, April, 2000.
Pietrobelli A, Faith M, Allison D, Gallagher D, Chiumello G &
Heymsfield S (1998): Body mass index as a measure of adiposity
among children and adolescent a validation study. J. Pediatr. 132,
204–210.
Poskitt E (1995): Defining childhood obesity: the relative body mass
index (BMI). Acta Pediatrica 84, 961–963.
Power C, Lake J & Cole T (1997): Measurement and long-term health
risks of child and adolescent fatness. Int. J. Obes. Relat. Metab.
Disord. 21, 507–526.
Reilly JJ, Dorosty AR, Emmett PM and the ALSPAC Study Team
(2000): Identification of the obese child: adequacy of the body
mass index for clinical practice and epidemiology. Int. J. Obes.
Relat. Metab. Disord. 24, 1623–1627.
Schroeder D & Martorell R (1999): Fatness and body mass index from
birth to young adulthood in rural Guatemalan population. Am. J.
Clin. Nutr. 70(Suppl): 137S–144S.
Troiano R & Flegal K (1998): Overweight and childhood and adoles-
cents: description, epidemiology and demographics. Pediatrics
101, 497–504.
Troiano R & Flegal K (1999): Overweight prevalence among youth in
the US: why so many different numbers. Int. J. Obes. Relat. Metab.
Disord. 23(Suppl 2): S22–S27.
Trowbridge FL, Marks JS, Lopez de Romana G, Madrid S et al (1987):
Body composition of Peruvian children with short stature and
high weight-height II. Implications for the interpretation for
weight for height as an indicator of nutritional status. Am. J.
Clin. Nutr. 46, 411–418.
Uauy R, Albala C & Kain J (2001): Obesity trends in Latin America:
transiting from under to over weight. J. Nutr. 131, 8935–8995.
Vio F & Albala C (2000): Nutrition policy in the Chilean transition.
Public Health Nutr. 3, 49–55.
World Health Organization Expert Committee on Physical Status
(1995): The use and interpretation of anthropometry. Report of a
WHO Expert Committee. WHO Technical Report Series no. 854
Geneva: WHO.
World Health Organization (1998) Obesity: Preventing and managing
the global epidemic. Report of a WHO Consultation on Obesity, 3–5
June 1997. Geneva: WHO.
Overweight and obesity in Chilean children
J Kain et al
204
European Journal of Clinical Nutrition