Obesity in Mexican children

1. Original Paper
Ann Nutr Metab 2008;52:227–232 Received: January 23, 2007
Accepted: October 4, 2007
DOI: 10.1159/000140514
Published online: June 19, 2008
Common Predictors of Excessive
Adiposity in Children from a Region
with High Prevalence of Overweight
Nancy Basaldúa Erwin Chiquete
Hospital Civil de Guadalajara ‘Fray Antonio Alcalde’, Centro Universitario de Ciencias de la Salud,
Universidad de Guadalajara, Guadalajara, México
Key Words having first-degree relatives with obesity (OR 2.59, 95% CI
Adiposity Body mass index Overweight children 1.41–4.74), female gender (OR 5.60, 95% CI 3.22–9.77) and be-
Obesity, Mexico Overweight, risk factors ing the third child or younger in offspring (OR 2.07, 95% CI
1.22–3.51). These effects could not be explained by social
class, ethnicity, maternal age and duration of breastfeeding.
Abstract Conclusions: Risk factors easily identified by history-taking
Aim: To identify risk factors other than energy intake or can predict childhood adiposity and the high risk of obesity
expenditure that can predict adiposity and overweight in in adulthood. Having a first-degree relative with obesity un-
children from a region with high prevalence of obesity. derscores the impact of genes and the family lifestyle on ex-
Methods: We studied 551 children aged 6–12 years (50.5% cessive adiposity. Being the third child or younger may de-
girls) from a city in the North of Mexico. Tetrapolar bioimped- note different nurture practices in offspring; however, this
ance was used to assess body fat content. Overweight was factor deserves more exploration.
estimated by analysis of age- and gender-standardized body Copyright © 2008 S. Karger AG, Basel
mass index (BMI) relative to reference data of the Interna-
tional Obesity Task Force (BMIs that predict obesity in adult-
hood). Multivariate analyses were modeled to find indepen- Introduction
dent predictors of adiposity. Results: The frequency of
overweight/obesity was 37.6%. There were no differences The prevalence of childhood overweight has increased
between genders with respect to weight, height and BMI; in most populations [1]. This picture is also observed in
however, age-standardized percentage of body fat and a low-income countries, where undernutrition is still a ma-
sedentary lifestyle were higher in girls than in boys (p ! jor issue [2–5]. These problems integrate the paradox of
0.001). Independent predictors of overweight/obesity were the low-income populations living the transition towards
having first-degree relatives with obesity [adjusted odds ra- a developed or ‘occidental’ lifestyle, in which both obe-
tio (OR) 2.26, 95% confidence interval (CI) 1.40–3.64], seden- sity and undernutrition concur affecting the less privi-
tary lifestyle (OR 1.58, 95% CI 1.05–2.37) and being the third leged classes [2, 6].
child or younger in offspring (OR 1.59, 95% CI 1.02–2.47). Pre- It is commonly thought that the medical illnesses as-
dictors of body fat in the highest quartile of the sample were sociated with obesity preferentially affect to adults; how-
© 2008 S. Karger AG, Basel Dr. Erwin Chiquete
0250–6807/08/0523–0227$24.50/0 Subdirección General de Enseñanza e Investigación
Fax +41 61 306 12 34 Hospital Civil de Guadalajara ‘Fray Antonio Alcalde’
E-Mail karger@karger.ch Accessible online at: Hospital 278, Guadalajara, Jalisco C.P. 44280 (México)
www.karger.com www.karger.com/anm Tel./Fax +52 33 3613 3951, E-Mail erwinchiquete@runbox.com

2. ever, in recent years it has been demonstrated that most ational physical activities), for at least 3 days per week. For an-
of the biological conditions that mediate the harmful thropometrics, children were measured in light sportwear after
they had emptied their bladders. Height was measured without
consequences of the excessive amount of adipose tissue shoes to the nearest 0.5 cm using a wall-mounted metric rule.
also occur early in life [7–11]. Childhood overweight pre- Waist and hip circumferences (cm) were measured with an an-
dicts obesity in the following decades [12, 13] and implies thropometric tape. Waist was measured at the minimum cir-
a high risk of cardiovascular disease and early mortality cumference between the iliac crest and the rib cage, below the
in adulthood [11]. Therefore, prevention and effective sternum. The hip measurement was taken at the maximum pos-
terior protrusion of the buttocks, around the greater trochanter.
treatment of overweight in children is essential to pro- Weight and body fat content were assessed by bioimpedance
mote a longer and healthier lifespan. analysis (BIA) with a four-pole impedance meter at 800 mAmp
There are several reports that have described deter- and 50 kHz (BIA 310 Bioimpedance Analyzer, Biodynamics, Se-
minants of childhood overweight, mainly lifestyle hab- attle, Wash., USA), at least 2 h after food ingestion. This instru-
its [14–16]. Nevertheless, little is known on whether ment has a maximum possible difference between any two mea-
surements for the same subject of 0.68 kg [95% confidence inter-
common early factors, other than caloric intake and en- val (CI) 0.58–0.84 kg] for estimation of body fat content in
ergy expenditure can predict adiposity in children per- children [21]. Children were asked to stand barefoot and without
taining to a population with a high prevalence of over- metals on an insulating sole, and electrodes were placed in the
weight. Therefore, we sought to identify independent four limbs, as corresponded. Gender and height details were en-
predictors of adiposity in childhood and of the high risk tered manually into the electronic system via a keyboard. Body
weight and total as well as percentage body fat (PBF) were esti-
of obesity in adulthood, in a cohort of schoolchildren mated using the standard built-in prediction algorithms for
from a region in the North of Mexico with one of the children. The printed report provided readings of fat mass, lean
highest prevalences of overweight in children and adults mass, bone mineral content, total body mass (sum of fat mass,
[17–19]. lean mass and bone mineral content), body mass index (BMI)
and PBF. The whole body scan time was 2–3 min. Overweight
was estimated by comparing BMI standardized for age and sex,
relative to reference data of the International Obesity Task Force
Materials and Methods (IOTF, standardized BMI cut-off points that predict overweight
and obesity at age 18 years) [13].
Study Population
This cross-sectional study was performed between February Statistical Analysis
and July 2004, in the urban population of Magdalena de Kino, The main dependent variables were PBF and the composite
State of Sonora. This city in the North of Mexico has a total pop- overweight/obesity, defined as the current BMI that predicts a fu-
ulation of about 24,447 inhabitants and is located at 80 km (49.6 ture BMI 625 when adult, according to IOTF reference tables
miles) from the frontier with the USA. The city of Magdalena de [13]. BMI and PBF were divided in percentiles, taking the highest
Kino has 33.5% of people aged !15 years [20], and pertain to a quartiles (percentile 75th or more) as dependent variables. Pear-
state of the Mexican Republic with an estimated prevalence of son 2 and Fisher exact tests were used to assess nominal variables
overweight and obesity combined of near 35%, in persons aged in bivariate and homogeneity analyses. To compare quantitative
3–17 years [18]. The internal Committee of Ethics of our institu- variables between two groups, Student’s t test and Mann-Whitney
tion approved the present study. Informed consent was obtained U test were performed in distributions of parametric and non-
from the children’s parents or legal proxy. parametric variables, respectively. Pearson correlation was used
in continuous variables (e.g. BMI, body fat content, height, weight,
Design and other somatometric variables). To find independent predic-
Parents and teachers of children from 2 public institutions of tors of adiposity, BMI in the highest quartile of the sample and of
the 16 elementary schools (either public or private) of the city were overweight/obesity, multivariate models were constructed by
asked for their alumni to participate in the analysis (669 children stepwise logistic regression. Input variables were those that re-
aged 6–13 years). Parents of 551 (82.3%) children responded to our sulted significantly associated with adiposity in bivariate analy-
request. A standardized, structured questionnaire was used to ses, but demographic variables and known risk factors for over-
collect data directly from the parents regarding demography, rel- weight (other than caloric intake) were also included in logistic
evant antecedents and current alimentary and exercise habits. regression analyses for adjustment, as potential confounders. Ad-
The questionnaire was administered as an interview by trained justed odds ratios with the respective 95% CIs are provided. The
personnel. Informed consent was obtained from the parents or fitness of the models was evaluated by using the Hosmer-Leme-
legal proxies. show goodness-of-fit test, which was considered as reliable if p 1
0.2. All p values are two-sided and considered significant when
Anthropometry and Assessment of Body Fat Content p ! 0.05. SPSS Version 13.0 for Windows (SPSS Inc., Chicago, Ill.,
A sedentary lifestyle was defined as 13 h per day spent sitting USA) was used for all statistical calculations.
down during leisure time (i.e., television watching, computer
use, and similar activities) in a child who is not engaged in a sys-
tematic exercise practice (i.e., sports, dance, and other recre-
228 Ann Nutr Metab 2008;52:227–232 Basaldúa/Chiquete

3. Table 1. Main characteristics of the 551 children analyzed
Variable Total Girls (n = 278) Boys (n = 273) p valuea
Age, median (range), years 9 (6–12) 9 (6–12) 9 (6–12) 0.58
Age, years 8.981.8 8.981.8 8.981.8 0.59
Child number 3 or more in offspring, n (%) 124 (28.9) 59 (26.6) 65 (31.4) 0.27
First-degree relative with obesity, n (%) 307 (71.7) 157 (69.8) 150 (73.9) 0.34
First-degree relative with diabetes, n (%) 225 (52.6) 120 (53.3) 105 (51.7) 0.74
Sedentary lifestyle, n (%)b 187 (43.7) 122 (54.2) 65 (32) <0.001
Height, m 1.3580.12 1.3580.13 1.3480.12 0.14
Weight, kg 35.1812.1 35.4812.2 34.9812.1 0.65
Waist circumference, cm 66.7810.5 66.2810.4 66.4810.6 0.51
BMI, kg/m2 18.784.0 18.683.9 18.984.1 0.51
BMI in the highest quartile, n (%) 138 (25) 69 (24.8) 69 (25.3) 0.90
PBF 17.487.3 20.486.4 14.487.0 <0.001
PBF in the highest quartile, n (%) 124 (24.7) 95 (37.5) 29 (11.6) <0.001
Overweight/obesity, n (%)c 207 (37.6) 106 (38.1) 101 (37) 0.78
Obesity, n (%)c 73 (13) 31 (11.2) 42 (15.4) 0.14
b
BMI = Body mass index; PBF = percentage of body fat. >3 h per day spent sitting down during leisure time (i.e., oth-
a p value for differences between boys and girls; Pearson 2 , er than school hours) and absence of systematic exercise practice
Student t test (for means) or Mann-Whitney U test (for medians), (i.e., sports, dance, and similar activities).
as corresponded. c
Current BMI that predicts overweight or obesity in adult-
hood, according to the International Obesity Task Force reference
tables.
Results
60
A total of 551 children aged 6–12 years were included Homogeneity, p = 0.02 *
in the final analysis (table 1). All the children were of Lat-
50
*
Frequency (%)
40
in-American ethnicity and their family had an annual * *
income of USD !15,000. 30 *
Age, ethnicity, social class, relevant antecedents, 20
height, weight, BMI, waist circumferences, the relative 10
frequency of obesity and the composite overweight/obe- 0
sity did not differ according to gender. However, seden- 6 7 8 9 10 11 12
Age (years)
tary lifestyle and a high body fat content were more fre-
quent in girls than in boys (table 1). Although the natural
increment of BMI with every year of age was identified,
Fig. 1. Distribution of the percentage of body fat in the highest
the relative frequency of overweight/obesity was homo- quartile across the age groups in girls (g), boys (k) and both
geneous across the age groups. PBF standardized as in- genders combined (i). * p ! 0.05.
crements of 1 year remained higher in girls than in boys
in age 6, and from 8 to 11 years (in all, p ! 0.01), but not
in age 7 and 12 years (fig. 1).
As expected, PBF positively correlated with height,
weight, BMI and waist circumference (fig. 2). Age also relative with obesity, sedentary lifestyle, and being the
moderately correlated with PBF (r = 0.174, r2 = 0.030, p ! third child or more in offspring. Independent predictors
0.001) and total body fat content (r = 0.426, r2 = 0.181, of PBF in the highest quartile were female gender, having
p ! 0.001). a first-degree relative with obesity and being the third
After multivariate analyses (table 2), independent pre- child or more in offspring. BMI in the highest quartile
dictors of overweight/obesity were having a first-degree was predicted only by the antecedent of a first-degree rel-
Predicting Excessive Adiposity in Ann Nutr Metab 2008;52:227–232 229
Children

4. 50 50
r = 0.224, r2 = 0.059, p < 0.001 r = 0.639, r2 = 0.408, p < 0.001
40 40
Body fat (%)
Body fat (%)
30 30
20 20
10 10
0 0
1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 10 20 30 40 50 60 70 80 90
a Height (m) b Weight (kg)
50 50
r = 0.791, r2 = 0.625, p < 0.001 r = 0.637, r2 = 0.406, p < 0.001
40 40
Body fat (%)
Body fat (%)
30 30
20 20
10 10
0 0
Fig. 2. Correlations between percentage 10 15 20 25 30 35 40 20 40 60 80 100 120
of body fat with height (a), weight (b), BMI c BMI (kg/m2) d Waist circumference (cm)
(c) and waist circumference (d). S = Girls,
U = boys.
Table 2. Determinants of excessive adiposity and overweight: three binary logistic regression models
Variables Multivariate odds ratios (95% CI)
overweight/obesitya PBF in the highest quartileb BMI in the highest quartilec
Female gender NS 5.60 (3.22–9.77) NS
First-degree relative with obesity 2.26 (1.40–3.64) 2.59 (1.41–4.74) 2.50 (1.40–4.44)
Sedentary lifestyle 1.58 (1.05–2.37) NS 1.95 (1.24–3.05)
Third child or more in offspring 1.59 (1.02–2.47) 2.07 (1.22–3.51) NS
b
BMI = Body mass index; PBF = percentage of body fat; NS = Hosmer-Lemeshow test for goodness-of-fit in the final step
non-significant variable, hence, not appearing as predictor after of the regression model: 2 = 5.67, 6 d.f., p = 0.46. Only significant
multivariate analysis, but included in prediction models as a con- predictors are shown. Adjusted for age, birth weight, duration of
founder. breastfeeding, maternal age, sedentary lifestyle and family his-
a
Hosmer-Lemeshow test for goodness-of-fit in the final step tory of diabetes mellitus.
of the regression model: 2 = 1.14, 6 d.f., p = 0.98. Only significant c
Hosmer-Lemeshow test for goodness-of-fit in the final step
predictors are shown. Adjusted for age, gender, birth weight, du- of the regression model: 2 = 0.03, 2 d.f., p = 0.99. Only significant
ration of breastfeeding, maternal age and family history of diabe- predictors are shown. Adjusted for age, gender, birth weight, du-
tes mellitus. ration of breastfeeding, maternal age, being the third child or
more in offspring and family history of diabetes mellitus.
230 Ann Nutr Metab 2008;52:227–232 Basaldúa/Chiquete

5. ative with obesity and a sedentary lifestyle. Therefore, the tions in weight [26]. However, with the methodology used
antecedent of obesity in family members was the con- in this study we could not account for other non-system-
stant predictor for the three measures of overweight. atic physical activities like playing outside or bicycle rid-
ing, as these activities were very inconstant in nature and
duration in the sample studied, but that can indeed affect
Discussion energy expenditure [26].
The constant predictor of the three measures of exces-
We found a high relative frequency of overweight/obe- sive adiposity was having a first-degree relative with obe-
sity. Over the past years, the prevalence of pediatric over- sity. This risk factor has been identified iteratively [11, 12,
weight has risen dramatically, so that 115% of the chil- 27] and underscores the impact of genes and a shared life-
dren are now considered overweight [22]. Our findings style on the accumulation of adipose tissue. The exact
are in agreement with previous reports about the preva- meaning of being the third child or younger in offspring
lence of overweight in Mexican children [17, 18]. Compa- and its relationship with overweight could not be deter-
rable dietary patterns are shared between persons living mined with the original methodology of our study. A
in the North of Mexico with those of southern USA [15, plausible explanation to this finding may be that with a
17], but, as is shown in the present report, with a higher large offspring the breastfeeding practice becomes more
frequency of overweight in the Mexican people, as com- difficult, leaving without this protective factor for child-
pared with US inhabitants [22]. It has been demonstrated hood overweight [28] to the younger offspring. This find-
that similar dietary habits in persons living in the same ing, however, deserves more exploration, since we did not
region, but with distinct backgrounds (i.e., race), are as- assess breastfeeding practices on patient’s siblings. An-
sociated with a higher frequency of overweight [23]. In a other deficiency that can be accounted on our study is the
previous study on 1,350 children from the North of Mex- method used to assess body fat content (i.e., bioimped-
ico, a 39% prevalence of overweight was found and a risk ance), which, although simple, inexpensive and accept-
factor related to this condition was the regular ‘crossing’ ably reliable when analyzing large cohorts [21], it is not
from Mexico to USA [17]. Different genes concurring in the most accurate method to estimate adiposity.
a similar environment imply different interactions and In summary, common variables included in a regular
consequences on health [24]. history-taking can predict childhood adiposity and the
In the present report, except for a more sedentary life- high risk of obesity in adulthood. In a population with
style and a higher body fat content in the female gender, high prevalence of obesity, the constant predictor of over-
there were no relevant differences between genders. Sed- weight in childhood is the family history of this condi-
entarism is a risk factor that is more common in girls than tion, which might underscore the importance of heritage
in boys [25] and together with the hormonal changes that or more likely the shared dietary and exercise habits.
characterize puberty, it may contribute to the higher body Identification of children at risk before they develop ex-
fat content observed in girls. We found that independent cessive adiposity is necessary to prevent unhealthy prac-
predictors of overweight/obesity were having a first-de- tices that led to a positive energy balance.
gree relative with obesity, a sedentary lifestyle, and being
the third child or younger in offspring. Factors associated
with PBF differed in that sedentarism was replaced by the Acknowledgments
female gender, as a predictor. On the other hand, BMI in
The authors are indebted to Dr. Sandra M. De la Herrán and
the highest quartile was associated with a sedentary life-
Dr. Martín A. Grijalva for their invaluable efforts and assistance
style and the family history of obesity. A concern may in anthropometry, interviews and important suggestions to this
arise on whether a sedentary lifestyle may be the conse- work. Also, the authors gratefully acknowledge the interest for
quence of excess weight rather than a cause. Indeed, with this study of the school authorities, directors and personnel of the
the cross-sectional design of this study we can only con- local health system in Magdalena de Kino, Sonora; as well as the
children’s parents.
clude that a lower physical activity is present in over-
weight children, but cannot assess whether the sedentary
lifestyle has preceded the weight gain. Recently it has
been demonstrated that variations in posture and simple
movements that are associated with the routines of daily
life could be biologically determined and precede varia-
Predicting Excessive Adiposity in Ann Nutr Metab 2008;52:227–232 231
Children

6. References
1 Ebbeling CB, Pawlak DB, Ludwig DS: Child- 10 Ovesen L: Adolescence: a critical period for 18 Hurtado-Valenzuela JG, Sotelo-Cruz N,
hood obesity: public-health crisis, common long-term tracking of risk for coronary heart Avilés-Rodríguez M, Peñuelas-Beltrán CI:
sense cure. Lancet 2002;360:473–482. disease? Ann Nutr Metab 2006;50:317–324. Aumento de la prevalencia de obesidad en
2 Caballero B: Obesity as a consequence of un- 11 Speiser PW, Rudolf MC, Anhalt H, Cama- niños y adolescentes de la consulta ambula-
dernutrition. J Pediatr 2006;149:S97–S99. cho-Hubner C, Chiarelli F, Eliakim A, Free- toria. Bol Clin Hosp Infant Edo Son 2005;22:
3 Villalpando S, Perez-Exposito AB, Shamah- mark M, Gruters A, Hershkovitz E, Iughetti 81–86.
Levy T, Rivera JA: Distribution of anemia as- L, Krude H, Latzer Y, Lustig RH, Pescovitz 19 National Survey on Health and Nutrition:
sociated with micronutrient deficiencies OH, Pinhas-Hamiel O, Rogol AD, Shalitin S, ENSANUT 2006. Instituto Nacional de
other than iron in a probabilistic sample of Sultan C, Stein D, Vardi P, Werther GA, Za- Salud Pública, México, 2006. Accessed Octo-
Mexican children. Ann Nutr Metab 2006;50: dik Z, Zuckerman-Levin N, Hochberg Z; ber 25, 2006 (www.insp.mx/ensanut).
506–511. Obesity Consensus Working Group: Con- 20 Statistical Information of the State of Sono-
4 Del Río-Navarro BE, Velazquez-Monroy O, sensus statement: Childhood obesity. J Clin ra. Gobierno del Estado de Sonora, México
Sanchez-Castillo CP, Lara-Esqueda A, Ber- Endocrinol Metab 2005;90:1871–1887. 2000. Accessed October 25, 2006 (http://
ber A, Fanghanel G, Violante R, Tapia-Cony- 12 Whitaker RC, Wright JA, Pepe MS, Seidel www.sonora.gob.mx/).
er R, James WP; Encuesta Nacional de Salud KD, Dietz WH: Predicting obesity in young 21 Sung RY, Lau P, Yu CW, Lam PK, Nelson EA:
2000 Working Group: National Health Sur- adulthood from childhood and parental Measurement of body fat using leg to leg bio-
vey 2000. Violante R, Tapia-Conyer R, James obesity. N Engl J Med 1997;337:869–873. impedance. Arch Dis Child 2001; 85: 263–
WP; Encuesta Nacional de Salud 2000 Work- 13 Cole TJ, Bellizzi MC, Flegall KM, Dietz WH: 267.
ing Group, National Health Survey 2000: Establishing a standard definition for child 22 Hedley AA, Ogden CL, Johnson CL, Carroll
The high prevalence of overweight and obe- overweight and obesity worldwide: interna- MD, Curtin LR, Flegal KM: Prevalence of
sity in Mexican children. Obes Res 2004;12: tional survey. BMJ 2000;320:1240–1243. overweight and obesity among US children,
215–223 [Erratum in: Obes Res 2004;12, fol- 14 Ludwig DS, Peterson KE, Gortmaker SL: Re- adolescents, and adults, 1999–2002. JAMA
lowing table of contents]. lation between consumption of sugar-sweet- 2004;291:2847–2850.
5 Raja’a YA, Bin Mohanna MA: Overweight ened drinks and childhood obesity: a pro- 23 Whitaker RC, Orzol SM: Obesity among US
and obesity among schoolchildren in Sana’a spective, observational analysis. Lancet urban preschool children: relationships to
City, Yemen. Ann Nutr Metab 2005;49:342– 2001;357:505–508. race, ethnicity, and socioeconomic status.
345. 15 Yngve A, Wolf A, Poortvliet E, Elmadfa I, Arch Pediatr Adolesc Med 2006; 160: 578–
6 Kain J, Vio F, Albala C: Childhood nutrition Brug J, Ehrenblad B, Franchini B, Haralds- 584.
in Chile: from deficit to excess. Nutr Res dottir J, Krolner R, Maes L, Perez-Rodrigo C, 24 Corthesy-Theulaz I, den Dunnen JT, Ferre P,
1998;18:1825–1837. Sjostrom M, Thorsdottir I, Klepp KI: Fruit Geurts JM, Muller M, van Belzen N, van
7 Mesa JL, Ortega FB, Ruiz JR, Castillo MJ, and vegetable intake in a sample of 11-year- Ommen B: Nutrigenomics: the impact of
Tresaco B, Carreno F, Moreno LA, Gutierrez old children in nine European countries: biomics technology on nutrition research.
A, Bueno M: Anthropometric determinants The Pro Children Cross-Sectional Survey. Ann Nutr Metab 2005;49:355–365.
of a clustering of lipid-related metabolic risk Ann Nutr Metab 2005;49:236–245. 25 Deheeger M, Bellisle F, Rolland-Cachera
factors in overweight and non-overweight 16 Reilly JJ, Armstrong J, Dorosty AR, Emmett MF: The French longitudinal study of growth
adolescents – influence of cardiorespiratory PM, Ness A, Rogers I, Steer C, Sherriff A; and nutrition: data in adolescent males and
fitness. The AVENA Study. Ann Nutr Metab Avon Longitudinal Study of Parents and females. J Hum Nutr Diet 2002;15:429–438.
2006;50:519–527. Children Study Team: Early life risk factors 26 Levine JA, Lanningham-Foster LM, Mc-
8 Moreno LA, Mesana MI, Fleta J, Ruiz JR, for obesity in childhood: cohort study. BMJ Crady SK, Krizan AC, Olson LR, Kane PH,
Gonzalez-Gross M, Sarria A, Marcos A, Bue- 2005;330:1357. Jensen MD, Clark MM: Interindividual vari-
no M; AVENA Study Group: Overweight, 17 Menchaca-Martín P, Zonana-Nacach A: ation in posture allocation: possible role in
obesity and body fat composition in Spanish Obesity in Mexican children from the north- human obesity. Science 2005;307:584–586.
adolescents. The AVENA Study. Ann Nutr ern border. Salud Publica Mex 2006;48:1. 27 Vogels N, Posthumus DL, Mariman EC,
Metab 2005;49:71–76. Bouwman F, Kester AD, Rump P, Hornstra
9 Flores-Calderon J, Gomez-Diaz RA, Rodri- G, Westerterp-Plantenga MS: Determinants
guez-Gomez G, Moran-Villota S: Frequency of overweight in a cohort of Dutch children.
of increased aminotransferases levels and Am J Clin Nutr 2006;84:717–724.
associated metabolic abnormalities in obese 28 Owen CG, Martin RM, Whincup PH, Dav-
and overweight children of an elementary ey-Smith G, Gillman MW, Cook DG: The ef-
school in Mexico City. Ann Hepatol 2005; 4: fect of breastfeeding on mean body mass in-
279–283. dex throughout life: a quantitative review of
published and unpublished observational
evidence. Am J Clin Nutr 2005; 82: 1298–
1307.
232 Ann Nutr Metab 2008;52:227–232 Basaldúa/Chiquete