Original Article

European Journal of Clinical Nutrition (2007) 61, 623–632. doi:10.1038/sj.ejcn.1602558; published online 29 November 2006

Overweight with concurrent stunting in very young children from rural Mexico: prevalence and associated factors

Guarantor: LC Fernald.

Contributors: LCF and LMN contributed to the design of the survey instruments, oversaw data collection, conducted statistical analysis, interpreted data and prepared the paper.

L C Fernald1 and L M Neufeld2

  1. 1Community Health and Human Development, School of Public Health, University of California, Berkeley, CA, USA
  2. 2Department of Nutritional Epidemiology, Instituto Nacional de Salud Pública (National Institute of Public Health), Cuernavaca, Morelos, Mexico

Correspondence: Dr LM Neufeld, Department of Nutritional Epidemiology, Instituto Nacional de Salud Pública (National Institute of Public Health), Cuernavaca, Morelos, Mexico. E-mail: neufeld@insp.mx

Received 27 January 2006; Revised 10 August 2006; Accepted 12 September 2006; Published online 29 November 2006.

Top

Abstract

Objective:

 

To document the prevalence of overweight or obesity concurrent with stunting in rural low-income Mexican children and to identify demographic and socio-economic characteristics that could help identify families at risk of having an overweight/obese and stunted young child in this population.

Design:

 

Cross-sectional analysis of the nutritional status of very young children, using primary data from a rural community-based survey conducted in 2003. Overweight, obesity and stunting were documented along with several maternal, household and community characteristics.

Setting:

 

Impoverished areas of rural Mexico.

Subjects:

 

Pre-school children (n=7555), aged 24–72 months.

Results:

 

The combined prevalence of overweight and obesity was equal to or greater than 20% in all children, as was the prevalence of stunting. The prevalence of concurrent overweight or obesity and stunting was approximately 5% in non-indigenous children, and over 10% in indigenous children 24–60 months. A multinomial logistic analysis revealed that the factors associated with coexisting stunting and overweight/obesity were lower socio-economic status (SES), lower maternal age, education, intelligence (vocabulary) and perceived social status, shorter maternal height, and larger household size. Among only stunted children, the risk of also being overweight or obese was associated with younger maternal age (relative risk ratios (RRR): 0.98, P=0.05), lower maternal perceived social status (RRR: 0.95, P<0.01) and maternal obesity (RRR: 2.93, P<0.0001) or overweight (RRR: 1.50, P=0.002).

Conclusions:

 

These analyses highlight that concurrent overweight or obesity and stunting is an important public health issue in low-income areas of rural Mexico beginning in early childhood. Even within this impoverished population, children living in households with low relative SES are the most vulnerable.

Sponsorship:

 

Financial support for this research was provided by the National Institutes of Child Health and Human Development, the Fogarty International Center at NIH, the John D and Catherine T MacArthur Foundation 'Research Network on Socioeconomic Status and Health' and the Mexican Government.

Keywords:

indigenous, poverty, socio-economic status, nutritional transition, pre-school, double burden

Top

Introduction

Mexico is currently in the midst of epidemiologic and nutritional transition, with a rapidly increasing prevalence of overweight and obesity occurring in regions also plagued by undernutrition (Barquera et al., 2006). Results from Mexico's National Nutrition Survey indicate that almost 20% of Mexican children aged 5–11 are currently classified as overweight or obese (Hernandez et al., 2003). The most recent National Health Survey, conducted in 2000, found obesity rates of 11–19% in children aged 10–17 (del Río-Navarro et al., 2004). Both studies showed higher prevalence in girls. Very little is known about the prevalence of obesity in younger, pre-school children in Mexico, although it is expected to be elevated given the patterns seen in older children. A high prevalence of obesity across childhood puts children at risk for numerous conditions, such as early-onset type II diabetes mellitus (Cruz et al., 2004).

Simultaneous with this increase in the occurrence of obesity and associated non-communicable chronic diseases, the national prevalence of growth retardation secondary to undernutrition (stunting) remains high. In Mexico, almost 20% of children under 5 years of age are stunted (Hernandez et al., 2003), and this value is much higher among specific subgroups, including rural (31.6%) and indigenous (44.3%) populations (Rivera and Sepúlveda-Amor, 2003). The functional consequences of stunting have been well documented (Martorell et al., 1998; Mendez and Adair, 1999; Grantham McGregor et al., 2000).

A handful of studies have reported the presence of obesity and growth retardation within the same children (Popkin et al., 1996; Mamabolo et al., 2005), although this phenomenon has not yet been documented in Mexico. There are clear implications for health professionals and policy makers who would benefit from understanding the extent of this problem. Thus, the first objective of this paper is to report the prevalence of overweight or obesity concurrent with stunting in rural low-income Mexican children from 2 to 6 years of age. A related objective is to document the prevalence of overweight or obesity (with or without stunting) in this younger population, which has not been studied previously.

The second major objective of this paper is to identify demographic, household and socio-economic characteristics that could help classify families at risk of having an overweight/obese, stunted young child in this low-income population. It is of paramount importance to know what factors could help to target vulnerable families because interventions designed to improve linear growth must not inadvertently contribute to excess weight accumulation.

Top

Research methods

Survey methods

Data were collected as part of Mexico's National Social Welfare Survey, which was conducted in 2003. Some of the communities participating in the survey had previously participated in a program impact evaluation (Oportunidades, previously Progresa), and others who had never received the program were visited for the first time in 2003. All communities were poor (income <20th percentile) and rural (defined as towns with <2500 inhabitants) from seven Mexican states (Guerrero, Hidalgo, Michoacán, Puebla, Querétaro, San Luis Potosí and Veracruz) (Figure 1). The sampling scheme used to identify the communities in the original survey and the identification of children within these communities have been described elsewhere (Behrman and Todd, 1999a, 1999b). Briefly, household eligibility was determined in two stages, first by identifying low-income communities and then by choosing low-income households within those communities (Behrman and Todd, 1999a, 1999b). Low-income communities were selected based on the proportion of households in those communities living in poverty using data from the 1995 National Census. Households were then selected according to an index of objective characteristics, such as housing materials, water and sanitation facilities, education, and family structure, which were shown to be good proxies for annual income. On average, 78% of the households in selected communities were classified as eligible for program benefits, and 97% of these households enrolled in the program (Behrman and Todd, 1999a, 1999b). Once enrolled, households received benefits for a minimum of 3 years conditional on meeting the program requirements; new households were not able to enroll until the next certification period. A similar scheme was used to identify new communities and households included in the 2003 survey. Communities were identified that had not previously been included in the program but had similar socio-economic and demographic characteristics. The mean daily per capita income in all communities was approximately $2US. All children identified between the ages of 2 and 6 years (24–72 months) were included in the current analysis, as were their mothers.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Map of Mexico with labels for seven states included in the survey.

Full figure and legend (43K)

Data collection and measures

Height and weight were measured for all children and their mothers by personnel trained and standardized according to international recommendations using standard techniques (Habicht, 1974; Lohman et al., 1989) and regularly calibrated portable scales (Tanita, HD-317 Digital Scale) and stadiometers (Seca 214, 'Road Rod'). Repeat measurements were taken from approximately 2% of the sample to monitor quality control, and all results were in the acceptable range of less than 5% variation. In accordance with current recommendations on the measurement of child overweight and obesity, body mass index (BMI) for age percentile was calculated. Overweight was defined as BMI for age percentile greater than or equal to the 85th but less than the 95th, and obesity was defined as BMI for age greater than or equal to the 95th percentile of the reference curve developed by the Centers for Disease Control (NCHS/CDC, 2005). Stunting was defined as having a Z-score for height-for-age (HAZ) less than or equal to -2 s.d. below the median (Hamill et al., 1977). It is well established that the potential influence of genetics on linear growth is minor compared to the impact of environmental factors such as nutritional status and infectious disease (WHO, 2006). In adults, BMI was calculated as weight in kilograms divided by the square of height in meters. Overweight in adults was defined as BMI 25.0–29.9 kg/m2 and obesity as BMI greater than or equal to 30.0 (WHO, 1995). Values of BMI greater than 60 and HAZ greater than 9.0 or less than -9.0 were excluded as implausible (<0.05% values were removed). Child's date of birth was validated using birth records.

Given the importance of maternal influence on pre-school weight and height in children, interviews were conducted with the mother or primary caregiver of each child. All instruments used in the survey were adapted to be sensitive to the extreme poverty in which the families were living, and all interviews were conducted in Spanish, or with a translator if the respondent's primary language was indigenous. The following information was obtained: age of the mother, education of the mother, number of people in the household, whether the father was present in the household, mother's marital status, mother's BMI, mother's height and whether an indigenous language was spoken at home. As a proxy for intelligence, a measure of working vocabulary of the caregiver was obtained using the Spanish language version of the Peabody Picture Vocabulary Test (PPVT-R), the Test de Vocabulario en Imagines Peabody (TVIP) (Dunn et al., 1986). The TVIP contains 125 items to assess vocabulary; the average score for the TVIP is set at 100, with an s.d. of 15. Items on the TVIP have been carefully selected through rigorous item analysis for their universality and appropriateness to Spanish-speaking communities, and have been used in low income, rural populations (Munoz et al., 1989; Umbel et al., 1992).

A proxy measure of socio-economic status (SES) was generated using household assets and housing quality. The housing and asset measurements were obtained in a baseline survey that occurred in 1997 (or by conducting a retrospective survey in the newly recruited households). Assets included: car, van, refrigerator, blender, television, gas heater, boiler, radio, stereo, video cassette recorder, washing machine and fan; housing characteristics included quality of roof, wall and floor, number of rooms, presence of indoor bathroom and presence of indoor electricity. These variables have been shown to provide good estimations of the economic concept of 'consumption', the gold standard measure of SES (Moser, 1998; McGee and Brock, 2001). Principal components analysis, a robust technique for reducing a large number of variables into one proxy measure, was used to summarize the variables into one measure (Montgomery et al., 2000; Falkingham and Namazie, 2002). The first principal component was retained under the assumption that the most common variation in the set of asset and housing variables would be a good proxy for household wealth (Filmer and Pritchett, 2001).

Subjective social status was measured in the mothers using the MacArthur Scale of Subjective Social Status (Ostrove et al., 2000). The ladders were included in this study in order to assess the contribution of self-reported social status over and above an objective measure of SES, such as education or assets. These ladders have been used in several studies with adults, and results suggest that ladder rankings are extremely powerful determinants of health-related outcomes (Adler et al., 2000; Goodman et al., 2001). The MacArthur Scale of Subjective Social Status asks participants to place themselves on a ladder in reference to the population. The instrument has two parts, one linked to traditional SES indicators (assessing placement based on income and education) and one linked to a more immediate, local environment (assessing placement in local community). The ladders are scaled from 0 (lowest) to 9 (highest); the mean value on the SES ladder in Taiwanese adults is 3.9 (plusminus1.9) (Hu et al., 2005), in Hungarian adults is 3.7–3.9 (plusminus0.02) (Kopp et al., 2004), and in adults in the UK and the US, the means are slightly higher (Ostrove et al., 2000; Singh-Manoux et al., 2005). Before using this measure in the population described here, focus groups and cognitive testing were conducted in order to ensure that the interpretation of the questions in a low-income Spanish-speaking population corresponded with the original intent of the questions in English. The two ladders were more highly inter-correlated than in previous studies (r=0.53, P<0.001), and for this reason, a summary score of the values on both ladders was calculated to use in subsequent analyses; thus the range of the variable was from 0 to 18.

A community survey was also conducted and assessed several conditions of the communities where the participants lived. Four key variables were selected for inclusion in the analyses described here, including (A) drainage (whether the community had a functional drainage system); (B) school breakfast program (whether the community was receiving the benefits of a federal breakfast program); (C) food package distribution program (whether the community was receiving 'despensas' or food packages including staples such as oil, rice beans etc.); (D) whether the majority of the community was indigenous.

Statistical methods

First, we generated descriptive statistics of the prevalence of stunting, overweight or obesity and concurrent stunting and overweight or obesity. Given that the prevalence of stunting is so high in indigenous populations in Mexico, prevalence estimates are presented by ethnicity (Rivera et al., 2003b).

We then conducted a multinomial logistic regression analysis in which the dependent variables were being stunted, overweight or obese or concurrently stunted and overweight or obese, and the comparison group was children who were neither stunted nor obese. The analysis was repeated using overweight and obese children as the comparison group to explore factors associated with being stunted, given the presence of overweight or obesity, and then using stunted children as the comparison group to look at factors associated with being obese, given being stunted. Relative risk ratios (RRR) and 95% confidence intervals are presented as results.

In all models, we adjusted for child sex, child age (as a dummy variable to allow for nonlinear associations), and maternal, household and community characteristics. Maternal characteristics, particularly BMI and height were included to reflect shared genes and environment. Other demographic and socio-economic variables tested in the analysis included household size and structure, family SES (maternal education, household assets and housing quality), parental marital status, maternal perceived social status, maternal intelligence (TVIP) and whether an indigenous language was spoken at home. Four community level variables, described above, were included as being indicative of community resources.

Statistical analyses were conducted using STATA 9.2 for Windows (STATA Corporation, College Station, TX, USA). Sampling weights were calculated to take into consideration unequal probabilities of selection resulting from sample design and non-response; survey commands were used for all analyses.

Ethical considerations

This study was approved by the Ethics Commission at the National Institute of Public Heath in Mexico, and the Committee on the Protection of Human Subjects at the University of California at Berkeley. Parents were invited to participate after receiving a detailed explanation of the survey procedures and they were then asked to sign an informed consent declaration; children were asked for their verbal assent after receiving a simple explanation of the procedures.

Top

Results

Sample

From the 10 797 households visited in the survey, a sample of 8241 children had complete height and weight data. Of these, 7555 had complete data sets available for the analyses reported here including all maternal, household and community level variables (sample sizes for each variable are shown in Table 1). There were no differences between the children for whom complete data were available and those for whom complete data were not available in terms of the prevalence of stunting, overweight or obesity.


Characteristics of mothers, households and communities

The mean age of the mothers was 31.4 (plusminus8.8) years (Table 1). They had a mean score on the TVIP of 82.2 (plusminus18.9), which is substantially lower than the standardized mean of 100. The majority of mothers had completed some formal schooling. The women were an average of 148.9 (plusminus5.7) cm tall, which is shorter than the national mean height of women 12–49 years of age of 152.9 cm (Rivera and Sepúlveda-Amor, 2003). The women had a mean BMI of 26.0 (plusminus4.5) kg/m2 (data not shown). Of the mothers, 17.1% could be classified as obese, and an additional 35.5% as overweight, which is equivalent to nationally representative data (Fernald et al., 2004). All families lived in conditions of poverty, with a large number of people per household, and over 30% of families were of indigenous origin. Functional drainage systems were present in less than 20% of communities, and almost three-quarters of the communities qualified for the federal breakfast program.

Characteristics of children

The mean age of children was 4.0 (plusminus1.1) years. Almost one-quarter of all children in this sample were overweight or obese; the prevalence did not differ according to ethnicity (Table 2). However, the prevalence of stunting in children from indigenous families was almost double that seen in those from non-indigenous families. For example, in 4- to 5-year-old children, the prevalence of stunting was 21.3% in non-indigenous children and 42.7% in indigenous children, which is similar to the difference between indigenous and non-indigenous children seen in a nationally representative sample (Rivera and Sepúlveda-Amor, 2003). The prevalence of concurrent obesity or overweight and stunting was also double in indigenous children. In 3-year-old children, for instance, the prevalence of stunting and overweight/obesity occurring in the same child was 5.9% in those from non-indigenous households and 12.1% in those from indigenous households.


Factors associated with stunting, overweight and obesity

Older children were more likely to be stunted than younger children, as were children with younger, shorter mothers, with less formal schooling and lower scores on the TVIP (Table 3). Lower SES, lower perceived social status, father's absence from the home, and a larger number of people in the household were also associated with being stunted.


Being overweight or obese was associated with the child being male, and having a young mother who was herself obese or overweight. A child with an obese mother had more than double the risk of being overweight or obese than a child without an obese mother; the risk to the child for having an overweight mother was also very high. Higher levels of household SES were associated with child overweight or obesity, whereas mother's education was not.

The factors associated with concurrent stunting and overweight/obese were lower maternal age, shorter maternal height, less maternal education, lower maternal intelligence (vocabulary), lower maternal perceived social status, lower SES and larger household size. When compared with children who were obese as the reference group, children who were obese and stunted came from bigger families (RRR: 1.18, P<0.001), had lower SES (RRR: 0.87, P=0.002), had mothers with lower education (RRR: 0.64, P<0.05 for contrast between highest and lowest tertile), lower social status (RRR: 0.95, P=0.002), lower maternal intelligence (vocabulary) (RRR: 0.98, P<0.001), lower maternal age (RRR: 0.97, P<0.001) and shorter maternal height (RRR: 0.87, P<0.001) (data not shown). In other words, the risk of stunting given that the child was obese was related primarily to indicators of low SES. Similarly, the risk of being stunted and obese, given being stunted was associated with lower maternal age (RRR: 0.98, P=0.05) and lower maternal perceived social status (RRR: 0.95, P<0.01). The risk of being overweight or obese given being already stunted was associated largely with maternal overweight (RRR: 1.50, P=0.002) and obesity (RRR: 2.93, P<0.001) (data not shown). Thus, when comparing children who are concurrently stunted and obese with children who are only stunted or obese, lower social status and lower maternal age increase the risk of concurrence, regardless of the group used for comparison.

Top

Discussion

This study provides evidence that stunting concurrent with overweight or obesity is an important public health problem, even among poor pre-school-aged children in rural Mexico. We have also shown that measures of low SES, such as larger family size, lower maternal education and lower maternal perceived social status, were strongly associated with the coexistence of stunting and overweight/obesity.

Approximately 20–25% of the 24- to 72-month-old children in this sample were overweight or obese, which is almost double the prevalence reported in the rural sub-sample of a nationally representative group of Mexican school-aged children, conducted 5 years previously (Hernandez et al., 2003). The prevalence of overweight and obesity that we report in the indigenous population was also higher than a more recent study of older indigenous children at boarding school (Monárrez-Espino et al., 2004). Although the use and validity of BMI in stunted populations has been questioned, recent evidence suggests that it continues to be the best field-appropriate method for estimating the risk of overweight and obesity in both stunted and non-stunted children (Hoffman et al., 2006).

We found a clear association between increased risk of overweight and obesity in children and several maternal characteristics, including increased maternal BMI. We have shown previously that there is a very high prevalence of overweight and obesity in the adults from this same rural, poor population (Fernald et al., 2004). Other research has identified parental overweight and obesity as the strongest determinant of childhood overweight (Strauss and Knight, 1999; Danielzik et al., 2004). Although genetic factors play a role in BMI and total body fat, evidence suggests that the shared environment is a much more important determinant of the high correlation between parental and child BMI (Franks et al., 2005).

A high proportion (at or over 20%) of children in this study were stunted, which was slightly higher than the national average for children under 5 years of age (17%) (Rivera and Sepúlveda-Amor, 2003). Consistent with previous reports (Rivera et al., 2003b; Zuniga et al., 2003; Avila-Curiel et al., 2004), the prevalence of linear growth retardation (stunting) in the indigenous children was almost double that of the national average. The prevalence of stunting did not vary over the 24–72 months age range, which is consistent with evidence that the critical period for the development of stunting is before 2 years of age, after which the prevalence and severity remain relatively constant, even if nutritional status is improved (Schroeder et al., 1995). Compared with children of height within the reference range, children who were stunted in this population had younger, shorter mothers, with less formal schooling and lower perceived social status; in addition, the stunted children were more likely to come from crowded homes, which is consistent with previous research (Menon et al., 2000; Larrea and Kawachi, 2005).

One in 10 indigenous children in the young population studied here were concurrently stunted and overweight or obese, whereas only one in 20 children were affected in non-indigenous populations. The relative proportion of stunted children who were also overweight or obese was similar among indigenous and non-indigenous groups, but the higher prevalence of stunting among indigenous children implies that a larger number suffer from concurrent stunting and overweight or obesity. Our analyses suggest that being indigenous, per se, was not the factor contributing to the increased risk of stunting or obesity. Instead, other variables reflective of low SES, such as lower maternal age, maternal working vocabulary and perceived social status and increased number of household members, contributed more significantly to the regression models. In Mexico, communities with largely indigenous populations are often more isolated, and have fewer community and health resources than non-indigenous populations. Furthermore, language barriers may limit the use of available resources as many individuals, particularly in the poor, rural communities in southern Mexico.

The simultaneous prevalence of overweight or obesity and linear growth retardation (stunting) has been noted in countries undergoing nutritional transition, such as Russia, Brazil and the Republic of South Africa (Popkin et al., 1996; Mamabolo et al., 2005), and there are some speculations as to the mechanisms at work. Some studies have suggested that mild stunting is a risk factor for the accumulation of excess body fat (Sawaya et al., 1998), particularly in girls (Mukuddem-Petersen and Kruger, 2004). There is also the possibility that diets are adequate or high in energy but low in essential micronutrients associated with linear growth, such as zinc (Rivera et al., 2003a). The dietary contribution of foods with high energy and low micronutrient density (e.g. simple carbohydrates, sugared beverages and fried snack foods) needs further clarification, particularly in the context of Latin America, where diets are changing rapidly (Popkin, 2001). It could be that intermittent periods of hunger without ready access to food cause children to overeat when food becomes available, as has been suggested in children with a history of malnutrition (Sawaya et al., 1998; González-Barranco and Rios-Torres, 2004). This situation could be exacerbated by the decreased ability for self-regulation seen in stunted children, which could contribute to opportunistic overeating (Hoffman et al., 2000). Lastly, populations in developing countries may be particularly vulnerable to lifestyle changes due either to undernutrition during fetal and early life and/or cycles of undernutrition throughout the life cycle (Neel et al., 1998). Under conditions of inadequate nutrient supply during intrauterine life, lifelong alterations in insulin sensitivity, energy metabolism and susceptibility to weight gain may be programmed (Hales and Barker, 2001). Owing to the cross-sectional nature of our data set, we are not able to comment on potential mechanisms at work in this population.

Our data show an increased risk of concurrent stunting and overweight/obesity associated with lower maternal perceived social status and increased family size. Even though all families who participated in the study were in the bottom 20th percentile of income distribution across Mexico, variability in economic status existed among the families. The significant association between perceived social status and increased risk for the co-occurrence of stunting and obesity/overweight in children suggests that even within a very poor population, there may be additional risk for families living at the lower end of the socio-economic spectrum. This trend has been observed in the US where children in low-income families are more likely to be overweight than children in families with high income (Alaimo and Olson, 2001), although this association has not been consistently replicated (Sobal and Stunkard, 1989).

Some potential limitations are apparent in the study described here. First, given that the sampling occurred as part of a national house-to-house survey with specific objectives, we could not ask extensive questions about dietary intake, physical activity patterns or other behaviors. Second, we do not have data about other factors that may be associated with overweight and obesity in this population, such as parenting style (Brewis, 2003), or television viewing (Dietz and Gortmaker, 2001), nor do we have data about household food insecurity. Future studies should use detailed interviews to assess possible covariates contributing to overweight and obesity in this poor rural population in which such information has not yet been adequately documented. Third, we are limited in our ability to make claims about causality, given that our data are cross-sectional.

Our data clearly point to a high prevalence of overweight and obesity in a very low income, rural Mexican population of pre-school children where the prevalence of linear growth stunting remains high. In addition, we show that one-tenth of indigenous children are concurrently stunted and overweight or obese, which is directly tied to low levels of SES. Given the link between childhood stunting and the increased risk of chronic disease in adulthood (Sawaya et al., 2003), the combination of stunting and overweight in early childhood that we report here is likely to increase even further the health risks for these children in later life (Grillol et al., 2005). Overweight and obesity are now major public health concerns in Mexico, even among pre-school children from very low-income families living in rural areas. Efforts to improve linear growth of young children must be designed cautiously so as not to further contribute to excessive weight gain. Longitudinal research on the determinants of linear growth faltering in populations with a high prevalence of overweight and obesity is urgently needed to ensure that interventions directed to vulnerable populations address these important public health problems simultaneously.

Top

References

  1. Adler NE, Epel ES, Castellazzo G, Ickovics JR (2000). Relationship of subjective and objective social status with psychological and physiological functioning: preliminary data in healthy white women. Health Psychol 19, 586–592. | Article | PubMed | ChemPort |
  2. Alaimo K, Olson CM, Frongillo Jr EA (2001). Low family income and food insufficiency in relation to overweight in US children: is there a paradox? Arch Pediatr Adolesc Med 155, 1161–1167. | PubMed | ChemPort |
  3. Avila-Curiel A, Shamah T, Barragán L, Chávez A, Avila MA, Juárez L (2004). Indice epidemiológico de nutrición infantil basado en un modelo polinomial de los valores de puntuación Z del peso para la edad [An epidemiological index to assess the nutritional status of children under five years of age in Mexico]. Arch Latinoam Nutr 54, 50–57. | PubMed | ChemPort |
  4. Barquera S, Hotz C, Rivera J, Tolentino L, Espinoza J, Campos I et al. (2006). Food consumption, food expenditure, anthropometric status and nutrition-related diseases in Mexico. Double Burden of Malnutrition in Developing Countries. FAO: Rome. pp 161–203.
  5. Behrman JR, Todd PE (1999a). Randomness in the experimental samples of PROGRESA (education, health and nutrition program). Available atwww.ifpri.org. International Food Policy Research Institute: Washington DC.
  6. Behrman JR, Todd PE (1999b). A report on the sample sizes used for the evaluation of the education, health and nutrition program (PROGRESA) of Mexico.Available atwww.ifpri.org. International Food Policy Research Institute: Washington DC.
  7. Brewis A (2003). Biocultural aspects of obesity in young Mexican schoolchildren. Am J Human Biol 15, 446–460. | Article |
  8. Cruz M, Torres M, Aguilar-Herrera B, Perez-Johnston R, Guzman-Juarez N, Aranda M.et al. (2004). Type 2 diabetes mellitus in children – an increasing health problem in Mexico. J Pediatr Endocrinol Metab 17, 183–190. | PubMed |
  9. Danielzik S, Czerwindki-Mast M, Langnäse K, Dilba B, Müller MJ (2004). Parental overweight, socioeconomic status and high birthweight are the major determinants of overweight and obesity in 5–7 y-old children: baseline data of the Kiel Obesity Prevention Study (KOPS). Int J Obes 28, 1494–1502. | Article | ChemPort |
  10. del Río-Navarro BE, Velázquez-Monroy O, Sánchez-Castillo CP, Lara-Esqueda A, Berber A, Fanghänel G et al. (2004). The high prevalence of overweight and obesity in Mexican children. Obes Res 12, 215–223. | PubMed |
  11. Dietz WH, Gortmaker SL (2001). Preventing obesity in children and adolescents. Annu Rev Public Health 22, 337–353. | Article | PubMed | ChemPort |
  12. Dunn LM, Padilla ER, Lugo DE, Dunn LM (1986). Test de Vocabulario en Imágenes Peabody (Peabody Picture Vocabulary Test): Adaptación Hispanoamericana (Hispanic-American Adaptation). American Guidance Service (Dunn Educational Services): Circle Pines.
  13. Falkingham J, Namazie C (2002). Measuring Health and Poverty: A Review of Approaches to Identifying the Poor. DFID (Department for International Development) Health Systems Resource Centre: London.
  14. Fernald LC, Gutierrez J-P, Neufeld LM, Mietus-Snyder M, Olaiz GO, Bertozzi SM et al. (2004). High prevalence of obesity among the poor in Mexico. JAMA 291, 2544–2545. | Article | PubMed | ChemPort |
  15. Filmer D, Pritchett LH (2001). Estimating wealth effects without expenditure data – or tears: an application to educational enrollments in states of India. Demography 38, 115–132. | Article | PubMed | ChemPort |
  16. Franks PW, Ravussin E, Hanson RL, Harper IT, Allison DB, Knowler WC et al. (2005). Habitual physical activity in children: the role of genes and the environment. Am J Clin Nutr 82, 901–908. | PubMed | ChemPort |
  17. González-Barranco J, Ríos-Torres JM (2004). Early malnutrition and metabolic abnormalities later in life. Nutr Rev 62, S134–S139. | Article | PubMed |
  18. Goodman E, Adler NE, Kawachi I, Frazier AL, Huang B, Colditz GA (2001). Adolescents' perceptions of social status: development and evaluation of a new indicator. Pediatrics 108, E31. | Article | PubMed | ChemPort |
  19. Grantham McGregor S, Fernald LC, Ani CC (2000). The role of nutrition in intellectual and behavioral development in children. In: Sternberg R (ed). Environmental Effects on Cognitive Abilities. Laurence Erlbaum Associates: Cambridge. pp 119–155.
  20. Grillol LP, Siqueira AF, Silva AC, Martins PA, Verreschi IT, Sawaya AL (2005). Lower resting metabolic rate and higher velocity of weight gain in a prospective study of stunted vs nonstunted girls living in the shantytowns of Sao Paulo, Brazil. Eur J Clin Nutr 59, 835–842. | Article | PubMed | ChemPort |
  21. Habicht JP (1974). Estandarización de métodos epidemiológicos cuantitativos sobre el terreno [Standardization of quantitative epidemiological methods in the field] (Article in Spanish). Bol Of Sanit Panam 76, 375–384. | ChemPort |
  22. Hales CN, Barker DJ (2001). The thrifty phenotype hypothesis. Br Med Bull 60, 5–20. | Article | PubMed | ISI | ChemPort |
  23. Hamill PV, Drizd TA, Johnson CL, Reed RB, Roche AF (1977). NCHS growth curves for children birth-18 years. United States. Vital Health Stat 11 (i-iv), 1–74.
  24. Hernandez B, Cuevas-Nasu L, Shamah-Levy T, Monterrubio E, Ramirez-Silva CI, Garcia-Feregrino R et al. (2003). Factors associated with overweight and obesity in Mexican school-age children: results from the National Nutrition Survey 1999. Salud Publica Mex 45 (Suppl 4), S551–S557. | PubMed |
  25. Hoffman DJ, Roberts SB, Verreschi I, Martins PA, de Nascimento C, Tucker KL et al. (2000). Regulation of energy intake may be impaired in nutritionally stunted children from the Shantitowns of Sao Paulo, Brazil. J Nutr 130, 2265–2270. | PubMed | ChemPort |
  26. Hoffman DJ, Sawaya AL, Martins PA, McCrory MA, Roberts SB (2006). Comparison of techniques to evaluate adiposity in stunted and nonstunted children. Pediatrics 117, e725–e732. | Article | PubMed |
  27. Hu P, Adler NE, Goldman N, Weinstein M, Seeman TE (2005). Relationship between subjective social status and measures of health in older Taiwanese persons. J Am Geriatr Soc 53, 483–488. | Article | PubMed |
  28. Kopp M, Skrabski A, Rethelyi J, Kawachi I, Adler NE (2004). Self-rated health, subjective social status, and middle-aged mortality in a changing society. Behav Med 30, 65–70. | PubMed |
  29. Larrea C, Kawachi I (2005). Does economic inequality affect child malnutrition? The case of Ecuador. Soc Sci Med 60, 165–178. | Article | PubMed |
  30. Lohman TG, Roche AF, Martorell R (1989). Anthropometric Standardization Reference Manual. Human Kinetics Books: Champaign, IL.
  31. Mamabolo RL, Alberts M, Steyn NP, Delemarre-van de Waal HA, Levitt NS (2005). Prevalence and determinants of stunting and overweight in 3-year-old black South African children residing in the Central Region of Limpopo Province, South Africa. Public Health Nutr 8, 501–508. | Article | PubMed |
  32. Martorell R, Ramakrishnan U, Schroeder DG, Melgar P, Neufeld L (1998). Intrauterine growth retardation, body size, body composition and physical performance in adolescence. Eur J Clin Nutr 52 (Suppl 1), S43–S52. discussion S52–S53. | PubMed |
  33. McGee R, Brock K (2001). From Poverty Assessment to Policy Change: Processes, Actors and Data. Institute of Development Studies: Brighton.
  34. Mendez M, Adair L (1999). Severity and timing of stunting in the first two years of life affect performance on cognitive tests in late childhood. J Nutr 129, 1555–1562. | PubMed | ISI | ChemPort |
  35. Menon P, Ruel MT, Morris SS (2000). Socio-economic differentials in child stunting are consistently larger in urban than in rural areas. Food Nutr Bull 21, 282–289.
  36. Monárrez-Espino J, Martínez H, Martínez V, Greiner T (2004). Nutritional status of indigenous children at boarding schools in northern Mexico. Eur J Clin Nutr 58, 532–540. | Article | PubMed | ChemPort |
  37. Montgomery MR, Gragnolati M, Burke KA, Paredes E (2000). Measuring living standards with proxy variables (in measurement and error). Demography 37, 155–174. | Article | PubMed | ChemPort |
  38. Moser CON (1998). The Asset Vulnerability Framework: Reassessing Urban Poverty Reduction Strategies. World Development 26, 1–19. | Article |
  39. Mukuddem-Petersen J, Kruger HS (2004). Association between stunting and overweight among 10-15-y-old children in the North West Province of South Africa: the THUSA BANA Study. Int J Obes Relat Metab Disord 28, 842–851. | Article | PubMed | ChemPort |
  40. Munoz F, Quilodran C, Velasquez P, Niedmann C, Baeza A, Silva G et al. (1989). Acquisition of the Spanish vocabulary among rural and urban students of the 9th region (Article in Spanish). Rev Child Pediatr 60, 354–358. | ChemPort |
  41. National Center for Health Statistics (2005). CDC Growth Charts, United States, Available at http://www.cdc.gov/growthcharts.
  42. Neel JV, Weder AB, Julius S (1998). Type II diabetes, essential hypertension, and obesity as 'syndromes of impaired genetic homeostasis': the 'thrifty genotype' hypothesis enters the 21st century. Perspect Biol Med 42, 44–74. | PubMed | ChemPort |
  43. Ostrove JM, Adler NE, Kuppermann M, Washingon AE (2000). Objective and subjective assessments of socioeconomic status and their relationship to self-rated health in an ethnically diverse sample of pregnant women. Health Psychol 19, 613–618. | Article | PubMed | ChemPort |
  44. Popkin BM (2001). The nutrition transition and obesity in the developing world. J Nutr 131, 871S–873S. | PubMed | ISI | ChemPort |
  45. Popkin BM, Richards MK, Montiero CA (1996). Stunting is associated with overweight in children of four nations that are undergoing the nutrition transition. J Nutr 126, 3009–3016. | PubMed | ChemPort |
  46. Rivera JA, Hotz C, Gonzalez-Cossio T, Neufeld L, Garcia-Guerra A (2003a). The effect of micronutrient deficiencies on child growth: a review of results from Community-Based Supplementation Trials. J Nutr 133, 4010S–44020. | ChemPort |
  47. Rivera JA, Monterrubio EA, González-Cossío T, García-Feregrino R, García-Guerra A, Sepúlveda-Amor J (2003b). Nutritional status of indigenous children younger than five years of age in Mexico: results of a national probabilistic survey. Salud Publica Mex 45, S466–S476. | PubMed |
  48. Rivera JA, Sepúlveda-Amor J (2003). Conclusions from the Mexican National Survey 1999: translating results into nutrition policy. Salud Publica Mex 45 (Suppl 4), S565–S575. | PubMed |
  49. Sawaya AL, Grillo LP, Verreschi I, da Silva AC, Roberts SB (1998). Mild stunting is associated with higher susceptibility to the effects of high fat diets: studies in a shantytown population in Sao Paulo, Brazil. J Nutr 128 (2 Suppl), 415S–420S. | PubMed | ChemPort |
  50. Sawaya AL, Martins P, Hoffman D, Roberts SB (2003). The link between childhood undernutrition and risk of chronic diseases in adulthood: a case study of Brazil. Nutr Rev 61 (5 Part 1), 168–175. | Article | PubMed |
  51. Schroeder DG, Martorell R, Rivera JA, Ruel MT, Habicht JP (1995). Age differences in the impact of nutritional supplementation on growth. J Nutr 125 (4 Suppl), 1051S–1059S. | PubMed | ChemPort |
  52. Singh-Manoux A, Marmot MG, Adler NE. (2005). Does subjective social status predict health and change in health status better than objective status? Psychosom Med 67, 855–861. | Article | PubMed |
  53. Sobal J, Stunkard AJ (1989). Socioeconomic status and obesity: a review of the literature. Psychol Bull 105, 260–275. | Article | PubMed | ISI | ChemPort |
  54. Strauss RS, Knight J (1999). Influence of the home environment on the development of obesity. Pediatrics 103, e85. | Article | PubMed | ChemPort |
  55. Umbel VM, Pearson BZ, Fernandez MC, Oller DK (1992). Measuring bilingual children's receptive vocabularies. Child Dev 63, 1012–1020. | Article | PubMed | ChemPort |
  56. WHO (1995). Physical Status: the use and interpretation of anthropometry. Technical Report Series, No. 854. World Health Organization: Geneva. pp 8–11.
  57. WHO (2006). The World Health Organization child growth standards. (http://www.who.int/childgrowth/standards/en/), accessed 1 August 2006.
  58. Zuniga MCC, Fritsch HM, Villa AR, Soto NG (2003). Alta prevalencia de denutrición en la población infantil indígena Mexicana. Encuesta nacional de nutrición 1999. [High prevalence of malnutrition among the indigenous early childhood population in Mexico. National Nutrition Survey 1999]. Rev Esp Salud Publica 77, 245–255. | PubMed |
Top

Acknowledgements

We acknowledge Aurora Franco, Ryo Shiba, Francisco Papaqui, Juan Pablo Gutierrez, Gustavo Olaiz and Stefano Bertozzi at the Instituto Nacional de Salud Publica, Paul Gertler and James Manley at the University of California, Berkeley, and Nancy Adler at the University of California, San Francisco. We thank the nurses who collected the data, and the families who participated in the study.

Top

MORE ARTICLES LIKE THIS

Extra navigation

.

natureevents

ADVERTISEMENT