SciELO - Scientific Electronic Library Online

vol.34 número2Lactancia materna para control del dolor agudo en lactantes: ensayo clínico controlado, ciego simpleEvaluación de indicadores antropométricos en niños con parálisis cerebral de acuerdo con el tipo de disfunción motora y estándar de referencia índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados

  • En proceso de indezaciónCitado por Google
  • No hay articulos similaresSimilares en SciELO
  • En proceso de indezaciónSimilares en Google


Nutrición Hospitalaria

versión On-line ISSN 1699-5198versión impresa ISSN 0212-1611

Nutr. Hosp. vol.34 no.2 Madrid mar./abr. 2017 

Trabajos originales

Factors associated with body mass index in Brazilian children: structural equation model

Factores asociados con el índice de masa corporal en niños brasileños: modelo de ecuaciones estructurales

Tatiane Géa-Horta1  , Rita de-C.-Ribeiro-Silva2  , Rosemeire L.-Fiaccone3  4  , Maurício L.-Barreto4  , Gustavo Velásquez-Meléndez1 

1Department Maternal Infant and Public Health. School of Nursing. Federal University of Minas Gerais. Belo Horizonte, Minas Gerais. Brazil.

2Nutrition Science Department. Nutrition School. Federal University of Bahia. Salvador, Bahia. Brazil.

3Department of Statistics. Institute of Mathematics. Federal University of Bahia. Salvador, Bahia. Brazil.

4Institute of Public Health. Federal University of Bahia. Salvador, Bahia. Brazil.



Overweight and obesity in children is an important global problem. Its prevalence is increasing in developed and developing countries.


The aim of this study was to evaluate the association between socioeconomic conditions, maternal body mass index (BMI), food security, and intake of obesogenic foods on excess body weight in children.


A cross-sectional study was conducted, using the data of 3,676 children from the latest National Demographic and Health Survey in Brazil. The children's BMI was the study outcome. Socioeconomic condition, maternal BMI, food security, and intake of obesogenic foods were use as predictors. structural equation models were used for analysis.


Socioeconomic conditions directly influenced the children's BMI ( = 0.102; p = 0.02), mediated by intake of obesogenic foods ( = 0.018; p = 0.04). A direct association was observed between maternal and child BMIs ( = 0.169; p < 0.001) and intake of obesogenic foods and child BMI ( = 0.114; p < 0.001).


Favorable socioeconomic conditions, increased maternal BMI, and intake of obesogenic foods contributed to increased child BMI.

Key words: Risk factors; Nutritional status; Child; Structural equation modeling; Brazil



el sobrepeso y la obesidad en los niños es un importante problema de salud global, habiéndose verificado aumento de la prevalencia en las poblaciones de los países desarrollados y en desarrollo.


el objetivo de este estudio fue estimar la asociación entre las condiciones socioeconómicas, el consumo y el exceso de peso corporal en los niños.


estudio transversal de una muestra probabilística representativa de la población brasileira que incluyó 3.676 niños provenientes de la última Encuesta Nacional de Demografía y Salud. El índice de masa corporal (IMC) de los niños fue la variable desenlace. El nivel socioeconómico, IMC materno, inseguridad a la hora de alimentarles y el consumo de alimentos obesogénicos fueron consideradas variables independientes. Las ecuaciones estructurales fueron usadas como método de análisis.


las condiciones socioeconómicas influyeron directamente el IMC de los niños ( = 0,102; p = 0,02), mediado por la ingesta de alimentos obesogénicos ( = 0,018; p = 0,04). Se observó asociación directa entre el IMC de la madre y del niño ( = 0,169; p < 0,001) y entre el consumo de alimentos obesogénicos y el IMC del niño ( = 0,114; p < 0,001).


la condición socioeconómica favorable, el IMC materno y la ingesta de alimentos obesogénicos pueden potencialmente contribuir para el aumento del IMC del niño.

Palabras clave: Factores de riesgo; Estado nutricional; Niños; Modelos de ecuaciones estructurales; Brasil


Overweight and obesity is an important global problem. Its prevalence is increasing in developed and developing countries 1,2, not only in adults, but also in children. This has been observed in North American 3, European 4, and Latin American 5 populations, including in Brazil 6. Its persistence throughout life can lead to several comorbidities 7 such as postural impairment, psychosocial problems 8, and metabolic disorders (dyslipidemia, arterial hypertension, insulin resistance, hyperinsulinemia, and diabetes) 9.

Genetic and postnatal factors (such as social, economic, cultural, psychosocial, and behavioral factors) are related to excess weight, making this a complex and multifaceted condition 8,10. The association between socioeconomic conditions and weight outcomes in children is controversial 2,11. Income and education are associated with food choices and food availability 12,13. Less economically privileged families are more susceptible to food and nutritional insecurity 12. Families with food insecurity adopt compensatory eating practices such as an increased consumption of low-cost, high-calorie, low-nutrient foods (14). Although several studies have indicated a positive association between food insecurity and excess weight 15,16, other studies found no such association 17,18.

Several studies have also indicated that parental, particularly maternal, overweight/obesity status influences excess weight in children, owing to genetic and family-behavioral components and life habits, including food intake and the unrestricted availability of high-calorie, low-nutrient foods at home 19,20.

Determinants of body weight gain in childhood have been investigated through models in which linear regression analysis, logistic analysis, and meta-analysis are commonly performed. However, few studies have applied structural equation modeling (SEM), which limits the simultaneous analysis of explanatory relationships between multiple variables, whether latent or observed.

In this context, the aim of this study was to evaluate the association between socioeconomic conditions, maternal body mass index (BMI), food security, and intake of obesogenic foods on excess body weight in children and how these interrelate to trigger this nutritional outcome. To evaluate these correlations, we used a theoretical model and consulted other studies as references 21.

For the present study, the following hypotheses were established a priori: a) increased child BMI is directly associated with unfavorable socioeconomic conditions, mediated by increased maternal BMI, food insecurity, and intake of obesogenic foods by the child, which are, in turn, directly associated with increased child BMI; b) increased child BMI is directly associated with increase in maternal BMI, mediated by the intake of obesogenic foods by the child; and c) increased child BMI is directly associated with food insecurity, mediated by the child's intake of obesogenic foods. Thus, we expected to obtain information that might support preventive and control programs against this nutritional phenomenon, particularly in children.


This cross-sectional study analyzed data from the latest National Demographic and Health Survey of Children and Women conducted between 2006 and 2007 in Brazil (PNDS) 22. The PNDS was designed to provide estimates representative of the Brazilian population living in private households (including slums), who were selected in 10 sampling strata that comprised a combination of all five major geographical Brazilian regions and both urban and rural areas. The sampling units were selected in two stages as follows: the primary units were composed of census tracts, and the secondary units were composed of households 22. Detailed sampling plans, data collection information and data quality assurance are available in the PNDS/DHS 2006 survey final reports at

Anthropometric measurements of the mothers and children were performed according to internationally standardized procedures 23, where two measures were obtained for each anthropometric indicator (weight and height) and the mean value was used. Mothers whose children were alive and living with them at the date of the interview were included in this analysis. If the mother had more than one child (approximately 22% of participants), the eldest child was chosen for this study. This choice was due to older children having had a longer duration of exposure to excess weight 24. The final number of child observations was 3,676. No statistically significant differences in socioeconomic and demographic characteristics were observed between the excluded and included children (data not shown). In addition, the amount of missing data was minimal (maternal education level, n = 44; maternal BMI, n = 16; household condition, n = 124; food insecurity, n = 109). Substitution of mean variable values was used for imputation of missing data 25.

BMI in children was expressed as a z-score in a continuous form and was considered the outcome variable. It was calculated as the weight (kg) divided by the square of the height (m2). The category excess weight, including overweight and obesity, measured by the z-score of the body mass index/age (BMIz) indicator was greater and equal than 1 z-scores. BMI z-scores were calculated using the standardized reference curves of the World Health Organization 26.

For the analysis via the SEM from the defined conceptual model, the latent variables considered and represented by an ellipse were as follows:

A. Intake of obesogenic foods by the child (foodobes): latent variable composed of the observed variables (represented by a rectangle): A.1. Intake of fried foods (friefood); A.2. Snacks (pretzels); and A.3. Sweets within the week preceding the interview. These three foods best represented the obesogenic foods. The frequency questionnaire of food consumption used to assess dietary intake of children consisted of a list of twenty foods, food groups, or preparations. The frequency of consumption was in reference to the seven days prior to the interview, and the following answers could be chosen: not consumed, consumed on one day, consumed on two or three days, consumed on four to six days, consumed every day, and do not know 27. To transform the categories of weekly frequencies into daily intake, a weight (Sn) was assigned to each food, in accordance with the intake frequency category as follows: Sn = (1/7)[(a + b)/2], where a and b correspond to the numbers of days of weekly intake frequency 27. A zero value was assigned to the "does not consume" or "rarely" frequency categories. The remaining categories of weekly frequency were the following: once, 2 or 3 times, 4-6 times, and daily (category to which a maximum weight of 1 was allocated).

B. Socioeconomic conditions(ses): latent variables composed of the following observed variables: B.1. Possession of household goods (hholdg), based on the economic classification criteria of Brazil of the Brazilian Association of Research Companies 28, considering the sum of goods, used as a continuous variable; B.2. Household conditions (housecon) and the presence (assigned the value of 0) or absence (assigned the value of 1) of piped water, linkage to a sewage network, electricity, and type of construction (masonry or other), were assessed. Then, a variable composed of the sum of the points of all items, ranging from zero (domicile that possessed all items) to three (domicile that had none of the items), was created; and B.3. Maternal education level (mateduc): years of study of the mother, used as a continuous variable.

The possible mediating observed variables were as follows:

A. Food insecurity (foodsec): assessed using the Brazilian Food Insecurity Scale, represented by scores (0: food security; 1: mild food insecurity; 2: moderate food insecurity; 3: severe food insecurity). This scale was adapted and validated for the non-institutionalized Brazilian population 29 and measures families' perceptions regarding food access. B. Maternal BMI (matbmi): calculated as weight (kg) over the square of height (m2), used as a continuous variable.

The SEM was used to assess the relationship between the study variables. Estimation of the direct and indirect effects of theoretical risk factors on child BMI provided more-realistic model tests and potentially greater statistical power than traditional multistep methods. In addition, under complex sampling, both point and variance estimators derived under independently and identically distributed observation assumptions are well known to potentially produce biased and inconsistent estimates. The model specified for SEM was built in Mplus 7.0 software by using robust maximum likelihood, where the strategy for replacing inverse Fisher information with a sandwich estimator of variance is useful for non-normality and non-independence of observations 30. To consider the complex survey design of the data, we used the "Type = Complex" statement, which included sampling weights and clustering of the PNDS data.

In relation of model fitness information, we choose to use relative (TLI- Tucker-Lewis Index) and the no centrality parameter (CFI- Bentler's Comparative Fit Index, RMSEA- Root Mean Square Error of Approximation) fit indices that are relatively unaffected by sample size. Values > 0.95 are suggested as cut-offs for these indices 31.

Regarding the ethical aspects, PNDS (2006) received the approval from the Ethics Research Committee (CEP) of the Center of Reference and Training DST/AIDS of the State Department of Health (SP). All individuals who agreed to participate in the study signed the informed consent form.


The mean age of the children was 2 years (standard deviation [SD] = 0.03), with similar frequency between sexes. The mean maternal age was 27 years (SD = 0.21), maternal BMI was 25 kg/m2 (SD = 0.16), and number of years of schooling was 8 years (SD = 0.11). Among the study children, 54.4% (95% confidence interval [CI] 0.51-0.58) lived in excellent housing conditions, 34.1% (95% CI 0.30-0.38) were among those whose possessions were within the first third of household goods, 6.4% (95% CI 0.05-0.08) had severe food insecurity, 4.9% (95% CI 0.04-0.06) consumed fried foods, 16.8% (95% CI 0.01-0.03) consumed snacks and 19.7% (95% CI 0.17-0.23) consumed sweets daily. The prevalence of excess weight, including overweight and obesity, was 7.3% (95% CI 0.06-0.09): 4.1% (95% CI 0.03-0.05) among boys and 3.2% (95% CI 0.03-0.04) among girls (data not shown).

Significant correlations were observed between certain variables and child BMI, such as maternal education (r = 0.05; p < 0.001), household goods (r = 0.11; p < 0.001), intake of fried foods (r = 0.05; p = 0.04), intake of sweets (r = 0.07; p = 0.01), and maternal BMI (r = 0.18; p < 0.001). Other correlations are shown in table I.

Table I Mean, standard deviation, and correlation coefficient of each study variable (Brazil, National Demographic and Health Survey of Children and Women, 2006-2007) 

*p < 0.001; **p < 0.05; BMI: body mass index; SD: standard deviation.

The factorial loads (FL) for the latent variable socioeconomic conditions ranged from -0.05 to 0.80, with the possession of household goods being the component that contributed most in this model of measurement (FL = 0.80), followed by maternal education (FL = 0.59; Fig. 1). The latent variable intake of obesogenic foods by the child (foodobes) was represented by the indicators of intakes of fried foods, snacks, and sweets within the week preceding the interview. The factorial loads ranged from 0.27 to 0.67, with the intake of snacks (FL = 0.67; Fig. 1) contributing most to this model, followed by the intake of sweets (FL = 0.59; Fig. 1).

Figure 1 Standardized estimates from structural equations modeling: measurement model and analysis of direct and indirect effects (note: hholdg: possession of household goods; mateduc: maternal education level; housecon: household conditions; ses: socioeconomic conditions; foodsec: food insecurity; matbmi: maternal body mass index; foodobes: intake of obesogenic foods by the child; childbmi: body mass index of child; fried food: intakes of fried foods). 

The direct and indirect effects (with their respective estimates) of socioeconomic conditions on child BMI, via maternal BMI, food security, and intake of obesogenic foods and how these variables are interrelated in triggering the outcome, are presented in table II and figure 1. We noted a direct association between socioeconomic conditions and child BMI ( = 0.102; p = 0.02) and an indirect effect mediated by the intake of obesogenic foods ( = 0.018; p = 0.04). This association was not mediated by maternal BMI ( = 0.006; p = 0.37) or food security ( = -0.030; p = 0.84). A direct positive association between socioeconomic conditions and intake of obesogenic foods was observed ( = 0.155; p = 0.02), along with an inverse association with food security ( = -0,544; p < 0.001). The data also indicated a direct association between maternal and child BMI ( = 0.169; p < 0.001) and intake of obesogenic foods ( = 0.114; p < 0.001). No other associations were statistically significant. The results of the SEM analysis attained acceptable values for adjustment indexes of the models (Table III).

Table II Direct and indirect effects of factors associated with child BMI (Brazil, National Demographic and Health Survey of Children and Women, 2006-2007) 

BMI: body mass index; : standardized beta coefficient; SD: standard deviation.

Table III Adjustments of the final model for the factors associated with child BMI (Brazil, National Demographic and Health Survey of Children and Women, 2006-2007) 

Socioeconomic conditions were observed to play an important role, affecting child BMI both directly (82.9% of the total effect; not mediated by other variables) and indirectly through maternal BMI (4.9% of the total effect) and the intake of obesogenic foods (14.6% of the total effect). The direct effect was 17 and 6 times greater than the indirect effect of maternal BMI and intake of obesogenic foods, respectively (Table IV). A similar effect was observed for maternal BMI (results not shown), although the indirect effect was not statistically significant.

Table IV Proportion of direct and indirect effects contributing to the total effect of socioeconomic conditions on child BMI (Brazil, National Demographic and Health Survey of Children and Women, 2006-2007) 

BMI: body mass index; CI: confidence interval; β: standardized beta coefficient.


From a nationally representative survey, the interrelations (direct and indirect) of socioeconomic conditions, maternal BMI, food security, and intake of obesogenic foods with the BMI z-scores of Brazilian children were tested using SEM.

Socioeconomic conditions positively influenced the increase in child BMI z-scores; this direct relationship (82.9%) was the most important association observed. It was further noted that a part of this total effect (14.6%) was mediated by the intake of obesogenic foods, indicating that as the socioeconomic level increases, food choices can lead to increased intake of unhealthy foods, which are predictors of increasing child BMI 32.

The association between an improved socioeconomic situation and unhealthy intake patterns may be indication of the changes in the modern lifestyle experienced by Brazilian families in the last few decades. In today's hectic world, the demand for practical and easy foods such as industrialized/processed foods, has increased. This increasingly contributes to the introduction of unhealthy intake patterns in families, including children's eating habits 33.

Results contrary to those observed in the present study showed that higher socioeconomic levels are associated with the adoption of consuming adequate and healthy foods 34. Evidence shows that higher maternal schooling and higher family income contribute positively to healthy food choices for children, as they allow parents to assimilate messages of nutritional education programs and understand the importance of diet as a way to promote health, leading to a reduced risk of excess weight in this age group 14. Nonetheless, the overestimation of the registry of healthy food intake, which occurs mainly among individuals with higher education levels, must not be disregarded. Having greater knowledge about healthy foods can lead them to exaggerate their report on their food intakes, thus concealing their actual food intakes 35. Therefore, this controversial correlation still needs to be clarified.

The findings of this study also revealed the direct and positive influence of obesogenic food intake on child BMI. Studies have analyzed the correlation between food intake patterns and excess weight, especially in children, and have shown that intake patterns based on junk food, characterized by a high intake of sweets, chocolate, ice cream, foods with added sugar, fried foods, and sugar-sweetened soft drinks, are related to excess weight. The effect of the intake of fatty foods on BMI gain is due to the high energy density associated with low levels of micronutrients that can result in an excessive passive intake, in which excess calories are unintentionally ingested 36,37. Most studies on this topic remain limited to cross-sectional studies, allowing investigation of the associations between variables only. However, several randomized and controlled intervention studies concluded that changes in eating practices to promote healthy eating positively influenced body weight reductions in all age groups 38.

Another important result of this study is the positive influence of maternal BMI on child BMI, reiterating findings from other studies that maternal obesity is one of the main predictors of excess weight in children 11,12. Although in the present study no direct positive association was observed between maternal BMI and the child's intake of obesogenic foods, the parents' influence on the development of their child's eating habits according to their preferences and attitudes toward diet could not be ignored, as it can interfere with the availability of food in households and can shape the child's nutritional status 22. One must also consider the role of genetics associated with obesity, which has been pointed out as an explanation for the occurrence of excess weight among individuals of the same family 10.

As expected, a reduction in food insecurity was directly associated with an increase in socioeconomic conditions. Despite the absence of statistical significance, food insecurity was negatively associated with child BMI, corroborating the results of another Brazilian study 17. However, our findings differ from those of studies in other underdeveloped and developing countries 39,40. In those studies, increased food insecurity was associated with increased body weight. According to the authors, the lack of financial resources often led families to adopt compensatory feeding practices, such as increasing the intake of high-calorie diets rather than foods with better nutritional quality. This may result in increased body weight in this segment of the population 16,17. These findings must be interpreted with caution, as socioeconomic status does not represent a direct or sufficient indicator of food security, because it does not take into account existing intra-family differences or specific problems such as maternal depression and parental dietary practices, which interfere with the food choices offered to the child 14.

As limitations, we highlight that the cross-sectional design of this study only allowed us to estimate associations between exposure and nutritional outcome; it cannot establish causal relationships. Another aspect that must be considered is that important factors associated with excess weight, such as children's physical activity level, hours of sleep, assisted use of television, and genetic factors were not considered in the models, as such information was not obtained by PNDS (2006-2007). On the other hand, the data from DHS are considered of high quality, as they follow strict standardization procedures and, in most cases, are the only sources of data on maternal and infant health to which several countries refer, especially those with developing or transitioning economies.

In conclusion, the results of this study confirm findings already reported in the literature, that favorable family socioeconomic conditions, increased maternal BMI, and intake of obesogenic foods contribute positively to increased child BMI. These data confirm the complexity of the interrelations between variables involved in determining child BMI and the need to mobilize several sectors of society to promote healthy habits and maintain healthy body weights in children.


TGH received a PhD scholarship from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES, Brazil and would also like to thank the Brazilian National Council of Research and Technology (CNPq) for a scholarship that supported a fellowship at the Federal University of Bahia, Brazil. The authors also thank FAPESB - Fundação de Amparo à Pesquisa do Estado da Bahia.


1. Okin BM, Adair LS, Ng SW. Global nutrition transition and the pandemic of obesity in developing countries. Nutr Rev 2012;70(1):3-21. [ Links ]

2. Wang Y, Lim H. The global childhood obesity epidemic and the association between socio-economic status and childhood obesity. Int Rev Psychiatry 2012;24(3):176-88. [ Links ]

3. Datar A, Chung PJ. Changes in socioeconomic, racial/ethnic, and sex disparities in childhood obesity at school entry in the United States. JAMA Pediatr 2015;169(7):696-7. [ Links ]

4. Bingham DD, Varela-Silva MI, Ferrao MM, Augusta G, Mourao MI, Nogueira H, et al. Socio-demographic and behavioral risk factors associated with the high prevalence of overweight and obesity in Portuguese children. Am J Hum Biol 2013;25(6):733-42. [ Links ]

5. Hernandez-Herrera RJ, Mathiew-Quiros A, Diaz-Sanchez O, Reyes-Trevino NO, Alvarez-Alvarez C, Villanueva-Montemayor D, et al. Prevalence of overweight and obesity in children from Monterrey, Nuevo Leon. Rev Med Inst Mex Seguro Soc 2014;52(Suppl 1):S42-S47. [ Links ]

6. Instituto Brasileiro de Geografia e Estatística (IBGE). Pesquisa de orçamentos familiares 2008-2009. Antropometria e estado nutricional de crianças, adolescentes e adultos no Brasil. Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística; 2010. [ Links ]

7. Sahoo K, Sahoo B, Choudhury AK, Sofi NY, Kumar R, Bhadoria AS. Childhood obesity: causes and consequences. J Family Med Prim Care 2015;4(2): 187-92. [ Links ]

8. Perez LM, Garcia K, Herrera R. Psychological, behavioral and familial factors in obese Cuban children and adolescents. MEDICC Rev 2013;15(4):24-8. [ Links ]

9. Sabin MA, Kiess W. Childhood obesity: current and novel approaches. Best Pract Res Clin Endocrinol Metab 2015;29(3):327-38. [ Links ]

10. Franks PW, Ling C. Epigenetics and obesity: the devil is in the details. BMC Med 2010;8:88. [ Links ]

11. Portela DS, Vieira TO, Matos SM, Oliveira NF, Vieira GO. Maternal obesity, environmental factors, cesarean delivery and breastfeeding as determinants of overweight and obesity in children: results from a cohort. BMC Pregnancy Childbirth 2015;15:94. [ Links ]

12. Parikka S, Maki P, Levalahti E, Lehtinen-Jacks S, Martelin T, Laatikainen T. Associations between parental BMI, socioeconomic factors, family structure and overweight in Finnish children: a path model approach. BMC Public Health 2015;15:271. [ Links ]

13. Bittencourt LS, Santos SMC, Pinto EJ, Aliaga MA, Ribeiro-Silva RC: Factors associated with food insecurity in households of public school students of Salvador City, Bahia, Brazil. J Health Popul Nutr 2013;31(4):471-9. [ Links ]

14. Eisenmann JC, Gundersen C, Lohman BJ, Garasky S, Stewart SD. Is food insecurity related to overweight and obesity in children and adolescents? A summary of studies, 1995-2009. Obes Rev 2011;12(5):e73-e83. [ Links ]

15. Rose D, Bodor JN. Household food insecurity and overweight status in young school children: results from the Early Childhood Longitudinal Study. Pediatrics 2006;117(2):464-73. [ Links ]

16. Jones SJ, Jahns L, Laraia BA, Haughton B. Lower risk of overweight in school-aged food insecure girls who participate in food assistance: results from the panel study of income dynamics child development supplement. Arch Pediatr Adolesc Med 2003;157(8):780-4. [ Links ]

17. Kac G, Schlussel MM, Perez-Escamilla R, Velasquez-Melendez G, Silva AA. Household food insecurity is not associated with BMI for age or weight for height among Brazilian children aged 0-60 months. PLoS One 2012;7(9):e45747. [ Links ]

18. Gundersen C, Garasky S, Lohman BJ. Food insecurity is not associated with childhood obesity as assessed using multiple measures of obesity. J Nutr 2009;139(6):1173-8. [ Links ]

19. Cebeci AN, Guven A. Does maternal obesity have an influence on feeding behavior of obese children? Minerva Pediatr 2014;67(6):481-7. [ Links ]

20. Novaes JF, Franceschini Sdo C, Priore SE. Mother's overweight, parents' constant limitation on the foods and frequent snack as risk factors for obesity among children in Brazil. Arch Latinoam Nutr 2008;58(3):256-64. [ Links ]

21. Kheirollahpour M, Shohaimi S. Dimensional model for estimating factors influencing childhood obesity: path analysis based modeling. Scientific World Journal 2014;2014:512148. [ Links ]

22. Brasil. Ministério da Saúde. Pesquisa Nacional de Demografia e Saúde da Criança e da Mulher - PNDS 2006: dimensões do processo reprodutivo e da saúde da criança. Brasília: Ministério da Saúde; 2009. [ Links ]

23. World Health Organization. Physical status: the use and interpretation of anthropometry. Geneva: World Health Organization; 1995. [ Links ]

24. Bartfeld J, Dunifon R. State-level predictors of food insecurity and hunger among households with children. United States Department of Agriculture; 2005. [ Links ]

25. Donders AR, Van der Heijden GJ, Stijnen T, Moons KG. Review: a gentle introduction to imputation of missing values. J Clin Epidemiol 2006;59(10):1087-91. [ Links ]

26. World Health Organization (WHO). Multicentre Growth Reference Study Group. Child Growth Standards: length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: methods and development. Geneva: World Health Organization; 2006. [ Links ]

27. Fornés NS, Martins IS, Velásquez-Meléndez G, Latorre MRDO. Escores de consumo alimentar e níveis lipêmicos em população de São Paulo, Brasil. Rev Saúde Pública 2002;36:12-8. [ Links ]

28. Associação Brasileira de Empresas de Pesquisas. Critério de Classificação Econômica do Brasil 2012. Dados com base no Levantamento Sócio Econômico 2010. Available at: [ Links ]

29. Perez-Escamilla R, Segall-Correa AM, Kurdian Maranha L, Sampaio Md Mde F, Marin-Leon L, Panigassi G. An adapted version of the U.S. Department of Agriculture Food Insecurity module is a valid tool for assessing household food insecurity in Campinas, Brazil. J Nutr 2004;134(8):1923-8. [ Links ]

30. Stapleton LM. An assessment of practical solutions for structural equation modeling with complex sample data. Struct Equ Modeling 2006;13:28-58. [ Links ]

31. Gerbing DW, Anderson JC. Monte Carlo evaluations of goodness-of-fit indices for structural equation models. In: Bollen KA, Long JS, editors. Testing structural equation models. Newbury Park, CA: Sage; 1993. [ Links ]

32. Silva RR, Marlúcia AOA, Szarfarc SC, Pinto EJ, Carneiro da Costa LC, Rodrigues LC. Iniquidades socioeconômicas na conformação dos padrões alimentares de crianças e adolescentes. Rev Nutr 2012;25:451-61. [ Links ]

33. Gwozdz W, Sousa-Pozab A, Reischa LA, Ahrensc W, Eibend G, Fernandéz-Alvirae JM, et al. Maternal employment and childhood obesity: a European perspective. J Health Econ 2013;32(4):728-42. [ Links ]

34. D'Innocenzo S, Marchioni DML, Prado MS, Matos SMA, Pereira SRS, Barros AP, et al. Condições socioeconômicas e padrões alimentares de crianças de 4 a 11 anos: estudo SCAALA - Salvador/ Bahia. Rev Bras Mater Health Infant 2011;11:41-9. [ Links ]

35. Hulshof KF, Brussaard JH, Kruizinga AG, Telman J, Lowik MR. Socio-economic status, dietary intake and 10 y trends: the Dutch National Food Consumption Survey. Eur J Clin Nutr 2003;57:128-37. [ Links ]

36. Santos NH, Fiaccone RL, Barreto ML, Silva LA, Silva R de C. Association between eating patterns and body mass index in a sample of children and adolescents in Northeastern Brazil. Cad Saude Pública 2014;30:2235-45. [ Links ]

37. Aranceta J, Perez-Rodrigo C, Ribas L, Serra-Majem L. Sociodemographic and lifestyle determinants of food patterns in Spanish children and adolescents: the enKid study. Eur J Clin Nutr 2003;57(Suppl 1):S40-S44. [ Links ]

38. De Bock F, Breitenstein L, Fischer JE. Positive impact of a pre-school-based nutritional intervention on children's fruit and vegetable intake: results of a cluster-randomized trial. Public Health Nutr 2011;15(3):466-75. [ Links ]

39. Hackett M, Melgar-Quinonez H, Alvarez MC. Household food insecurity associated with stunting and underweight among preschool children in Antioquia, Colombia. Rev Panam Salud Publica 2009;25(6):506-10. [ Links ]

40. Dubois L, Francis D, Burnier D, Tatone-Tokuda F, Girard M, Gordon-Strachan G, et al. Household food insecurity and childhood overweight in Jamaica and Quebec: a gender-based analysis. BMC Public Health 2011;11:199. [ Links ]

Received: July 11, 2016; Accepted: December 13, 2016

Correspondence: Gustavo Velásquez-Meléndez. Department Maternal Infant and Public Health. School of Nursing. Federal University of Minas Gerais. Belo Horizonte, Minas Gerais. Brazil e-mail:

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY-NC-SA 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.