ABSTRACT

OBJECTIVE

To analyze the dietary patterns of Brazilian children under two years of age and assess their association with sociodemographic characteristics and health service use.

METHODS

This is a cross-sectional study with data from the 2013 National Health Survey (PNS). Patterns were found for two age groups by principal component analysis and their correlation with characteristics of interest was tested by linear regression models.

RESULTS

We found two dietary patterns for our groups. The first consisted of the consumption of fresh or minimally processed foods and the second, of ultra-processed foods. The greater adherence of children between six and 11 months to the first pattern was associated with higher per capita family income and urban residences in the most developed regions of Brazil. At 12 months or more, adherence related to white race/color, higher per capita family incomes, residence in more developed regions, and visits to private childcare. Adherence to the second pattern among children under one year of age was inversely associated with Yellow or Indigenous race/color, residence in the Brazilian Northeast, and childcare in specialized public or private services. At 12 months or more, greater adherence was directly associated with Black or Brown children who resided in more developed regions, and inversely associated with those living in the Brazilian Northeast.

CONCLUSION

We found two opposite dietary patterns in Brazilian children under two years of age and that several social determinants modify their chance of adhering to these patterns.

Infant Nutrition; Diet, Food, and Nutrition; Socioeconomic Factors; Child Health Services

INTRODUCTION

Children’s first two years of life are crucial for their growth and development. At this stage, adequate and healthy eating plays a fundamental role for the satisfactory occurrence of these processes¹1. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Promoção da Saúde. Guia alimentar para crianças menores de dois anos. Brasília, DF; 2019.. As long as it is offered in appropriate quantities and quality, food is associated with the best health conditions in childhood and can prevent several diseases which develop in adulthood¹1. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Promoção da Saúde. Guia alimentar para crianças menores de dois anos. Brasília, DF; 2019..

On the other hand, inadequate nutrition is the most important risk factor for the loss of life years and is associated with several highly prevalent chronic non-communicable diseases, such as obesity, type 2 diabetes, hypertension, and some types of cancer²2. GBD 2015 Risk Factors Collaborators. Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016;388(10053):1659-724. https://doi.org/10.1016/S0140-6736(16)31679-8
https://doi.org/10.1016/S0140-6736(16)31... . Moreover, it is one of the main determinants of the global syndemic of obesity, malnutrition, and climate change, affecting most people in the world³3. Swinburn B, Kraak VI, Allender S, Atkins VJ, Baker PI, Bogard JR, et al. The Global syndemic of obesity, undernutrition, and climate change: the Lancet Commission report. Lancet. 2019;393(10173):791-846. https://doi.org/10.1016/S0140-6736(18)32822-8
https://doi.org/10.1016/S0140-6736(18)32... . These consequences are particularly relevant for children, since health behaviors and damages acquired in the early stages of life tend to have repercussions at later ages⁴4. Turer CB, Brady TM, Ferranti SD. Obesity, hypertension, and dyslipidemia in childhood are key modifiable antecedents of adult cardiovascular disease. a call to action. Circulation. 2018;137(12):1256-9. https://doi.org/10.1161%2FCIRCULATIONAHA.118.032531
https://doi.org/10.1161%2FCIRCULATIONAHA... .

Previous studies show that substituting the consumption of healthy meals (based on fresh or minimally processed foods) by ready-to-eat or semi-ready ones (i.e., ultra-processed) constitute one of the main determinants of worsening food quality⁵5. Rauber F, Costa Louzada ML, Steele EM, Millett C, Monteiro CA, Levy RB. Ultra-processed food consumption and chronic non-communicable diseases-related nutrient profile in the UK (2008-2014). Nutrients. 2018;10(5):587. https://doi.org/10.3390/nu10050587
https://doi.org/10.3390/nu10050587... ,⁶6. Costa Louzada ML, Ricardo CZ, Steele EM, Levy RB, Cannon G, Monteiro CA. The share of ultra-processed foods determines the overall nutritional quality of diets in Brazil. Public Health Nutr. 2018;21:94-102. https://doi.org/10.1017/s1368980017001434
https://doi.org/10.1017/s136898001700143... . Children’s consumption of unhealthy foods has not only grown but seems also to be associated with a higher risk of developing obesity⁷7. Costa CS, Rauber F, Leffa PS, Sangalli CN, Campagnolo PDB, Vitolo MR. Ultra-processed food consumption and its effects on anthropometric and glucose profile: a longitudinal study during childhood. Nutr Metab Cardiovasc Dis. 2019;29(2):177-84. https://doi.org/10.1016/j.numecd.2018.11.003
https://doi.org/10.1016/j.numecd.2018.11... and dyslipidemia⁸8. Rauber F, Campagnolo P, Hoffman D, Vitolo M. Consumption of ultra-processed food products and its effects on children’s lipid profiles: a longitudinal study. Nutr Metab Cardiovasc Dis. 2015;25(1):116-22. https://doi.org/10.1016/j.numecd.2014.08.001
https://doi.org/10.1016/j.numecd.2014.08... ; the former is the most common metabolic factor in the development of other chronic diseases among children, such as diabetes and high blood pressure⁴4. Turer CB, Brady TM, Ferranti SD. Obesity, hypertension, and dyslipidemia in childhood are key modifiable antecedents of adult cardiovascular disease. a call to action. Circulation. 2018;137(12):1256-9. https://doi.org/10.1161%2FCIRCULATIONAHA.118.032531
https://doi.org/10.1161%2FCIRCULATIONAHA... ,⁹9. Bhupathiraju SN, Hu FB. Epidemiology of obesity and diabetes and their cardiovascular complications. Circ Res. 2016;118:1723-35. https://doi.org/10.1161/circresaha.115.306825
https://doi.org/10.1161/circresaha.115.3... .

These findings are especially important since children’s first years of life include not only their introduction to complementary foods which are essential to meet their nutritional and developmental needs but also the possibility of establishing parental practices, preferences, and eating habits (whether healthy or not¹⁰10. Silva GAP, Costa KAO, Giugliani ERJ. Infant feeding: beyond the nutritional aspects. J Pediatr (Rio J). 2016;923(3 Suppl 1):S2-7. https://doi.org/10.1016/j.jped.2016.02.006
https://doi.org/10.1016/j.jped.2016.02.0... ), which also requires the continuous evaluation of consumption markers in this age group.

Moreover, only few studies have evaluated the influence of health service use on infant feeding. In general, studies which evaluated mothers who used this support have focused on identifying aspects related to the monitoring and performance of indicators, especially of breastfeeding¹¹11. Fernandes RC, Höfelmann DA. Intenção de amamentar entre gestantes: associação com trabalho, fumo e experiência prévia de amamentação. Cienc Saude Colet. 2020;25(3):1061-1072. https://doi.org/10.1590/1413-81232020253.27922017
https://doi.org/10.1590/1413-81232020253... ,¹²12. Venancio SI, Giugliani ERJ, Silva OLO, Stefanello J, Benicio MHDA, Reis MCG, et al. Associação entre o grau de implantação da Rede Amamenta Brasil e indicadores de amamentação. Cad Saude Publica. 2016;32(3):e00010315. https://doi.org/10.1590/0102-311X00010315
https://doi.org/10.1590/0102-311X0001031... , and the existence of a link between users and healthcare providers¹³13. Reichert APS, Rodrigues PF, Albuquerque TM, Collet N, Minayo MCS. Vínculo entre enfermeiros e mães de crianças menores de dois anos: percepção de enfermeiros. Cienc Saude Colet. 2016;21(8):2375-82. https://doi.org/10.1590/1413-81232015218.07662016
https://doi.org/10.1590/1413-81232015218... . Moreover, they claim that introducing unrecommended foods in children’s first year of life is associated with healthcare providers failing to effectively guide parents and guardians on nutrition and the latter’s low adherence to implemented nutritional prevention and education actions, especially in primary care services¹⁴14. Dallazen C, Silva SA, Gonçalves VSS, Nilson EAF, Crispim SP, Lang RMF, et al. Introdução de alimentos não recomendados no primeiro ano de vida e fatores associados em crianças de baixo nível socioeconômico. Cad Saude Publica. 2018;34(2):e00202816. https://doi.org/10.1590/0102-311x00202816
https://doi.org/10.1590/0102-311x0020281... . This scenario also requires further investigations.

Considering that the analysis of the simultaneous effect of different food groups¹⁵15. Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol. 2002;13(1):3-9. https://doi.org/10.1097/00041433-200202000-00002
https://doi.org/10.1097/00041433-2002020... tends to better represent food intake, previous studies conducted in schoolchildren or residents of certain Brazilian regions evaluated food intake by assessing dietary patterns, observing that maternal characteristics tend to influence children’s adherence to these patterns¹⁶16. Bogea EG, Martins MLB, Carvalho WRC, Arruda SPM, França AKTC, Silva AAM. Padrões alimentares de crianças de 13 a 35 meses de idade e associação com características maternas. Cad Saude Publica. 2019;35(4):e00072618. https://doi.org/10.1590/0102-311x00072618
https://doi.org/10.1590/0102-311x0007261... ,¹⁷17. Sotero AM, Cabral PC, Silva GAP. Socioeconomic, cultural and demographic maternal factors associated with dietary patterns of infants. Rev Paul Pediatr. 2015;33(4):445-52. https://doi.org/10.1016/j.rpped.2015.03.006
https://doi.org/10.1016/j.rpped.2015.03.... . However, these findings neither enable the generalization of the dietary patterns and sociodemographic factors associated with the entire national territory and all age groups in the infant population nor provide evidence of this population’s dietary patterns which may be associated with health service use. This gap enables us to conduct new studies which analyze information in a nationally representative sample of Brazilian children, such as those under two years of age.

Thus, based on population sample data related to infant feeding, made available by the National Survey of Health in 2013¹⁸18. Instituto Brasileiro de Geografia e Estatística, Diretoria de Pesquisas, Coordenação de Trabalho e Rendimento. Pesquisa Nacional de Saúde 2013: percepção do estado de saúde, estilos de vida e doenças crônicas. Rio de Janeiro: IBGE; 2014., and robust exploratory analysis methods (which would enable us to assess and explain the most frequent food consumption combinations), this study aims to analyze the dietary patterns of Brazilian children under two years of age and evaluate their association with sociodemographic characteristics and health service use.

METHODS

Design, Study Population, and Data Collection

This is a cross-sectional, household-based study with secondary data from the Brazilian National Health Survey (PNS), conducted by the Instituto Brasileiro de Geografia e Estatística (IBGE – Brazilian Institute of Geography and Statistics), in partnership with the Ministry of Health. Data were collected by trained interviewers from August 2013 to February 2014.

PNS used a three-stage probabilistic sampling: first, primary sampling units were composed of census tracts; then, households were drawn; and finally, residents aged 18 years or older. At each stage, selection was conducted by simple random sampling. PNS census tracts or sets were drawn via a master sample based on IBGE Integrated System of Household Surveys, with greater geographic spread and estimate precision. Thus, PNS represents Brazil, its macro-regions, federation units, metropolitan regions, and capitals.

Interview records were obtained from 60,202 households (an 86.1% response rate). Children under two years of age were identified in the PNS second stage, in which a fixed number of permanent private households was chosen from each primary sampling unit¹⁹19. Souza-Júnior PRB, Freitas MPS, Antonaci GA, Szwarcwald CL. Desenho da amostra da Pesquisa Nacional de Saúde 2013. Epidemiol Serv Saude. 2015;24(2):207-16. https://doi.org/10.5123/S1679-49742015000200003
https://doi.org/10.5123/S1679-4974201500... . Mothers or guardians answered the questions regarding the children in the chosen households. Only 3,646 children between six and 23 months of age were included in the final sample of this study.

Sampling weights were defined for the primary sampling units, households, and all their residents. Additional details about the sampling process, weighting factors, collection, and other information can be found in a previous publication¹⁹19. Souza-Júnior PRB, Freitas MPS, Antonaci GA, Szwarcwald CL. Desenho da amostra da Pesquisa Nacional de Saúde 2013. Epidemiol Serv Saude. 2015;24(2):207-16. https://doi.org/10.5123/S1679-49742015000200003
https://doi.org/10.5123/S1679-4974201500... .

Study Variables

Food intake, sociodemographic characteristics, and health service use data were analyzed.

Food Intake

Food intake was assessed by a questionnaire, applied to children’s guardians, aiming to evaluate the consumption of the following foods in the 24 hours prior to data collection: non-breast milk or dairy products; fruits or natural juices; vegetables; beans or other legumes (lentils, peas, etc.); meat or eggs; potato and other tubers and roots (sweet potato, cassava); cereals and derivatives (rice, bread, cereal, pasta, flour, etc.); artificial juices; cookies or cake; sweets or other foods with sugar; and soft drinks. All these variables were categorized into dichotomous indicators (0 = did not consume; 1 = consumed).

Sociodemographic Characteristics

The following sociodemographic variables were considered: sex (male; female); age (6 to 11 months and 29 days; 12 to 23 months and 29 days); race/color (white; Black or Brown; other – Yellow or Indigenous); per capita family income (distributed in quintiles); area of residence (rural; urban); and major regions (North; Northeast; Southeast; South; Midwest).

Health Service Use Characteristics

The following health service use variables were included: health insurance possession (yes; no) and place of children’s growth and development monitoring (childcare) (no follow up; basic health unit; specialized public service - specialty center, public polyclinic or health center, public hospital/outpatient clinic; private clinic; and others).

Data Analysis

Sociodemographic distribution, health service use, and the prevalence of the food consumed in our sample were shown as percentages and their respective 95% confidence intervals (95%CI).

Food intake patterns were defined via factor analysis by principal components¹⁵15. Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol. 2002;13(1):3-9. https://doi.org/10.1097/00041433-200202000-00002
https://doi.org/10.1097/00041433-2002020... , a method indicated when data show abnormal distribution²⁰20. Costello AB, Osborne J. Best practices in exploratory factor analysis: four recommendations for getting the most from your analysis. Pract Assess Res Eval. 2005;10:173-8. https://doi.org/10.7275/jyj1-4868
https://doi.org/10.7275/jyj1-4868... . Then, the obtained factors were rotated by orthogonal varimax rotation. The Bartlett and Kayser-Meyer-Olkin’s (KMO) tests of sphericity were applied to assess data adequacy for factor analysis²¹21. Gleason PM, Boushey CJ, Harris JE, Zoellner J. Publishing nutrition research: a review of multivariate techniques – part 3: data reduction methods. J Acad Nutr Diet. 2015;115:1072-82. https://doi.org/10.1016/j.jand.2015.03.011
https://doi.org/10.1016/j.jand.2015.03.0... . The factors which met an eigenvalue > 1 and scree plot graphical analysis criteria were maintained. Best factor interpretability was used to compose final components and foods with factor loadings equal to or greater than 0.3 were considered important for finding dietary patterns. Foods with cross-loadings – equal to or greater than the specified load value (0.3) in two or more factors – were removed from analysis to identify and isolate data which best fit our factorial structure²⁰20. Costello AB, Osborne J. Best practices in exploratory factor analysis: four recommendations for getting the most from your analysis. Pract Assess Res Eval. 2005;10:173-8. https://doi.org/10.7275/jyj1-4868
https://doi.org/10.7275/jyj1-4868... . Then, factorial scores were predicted for each evaluated child.

Associations between predicted scores (dependent variable), sociodemographic variables, and health service use (independent) were evaluated by linear regression models. In our adjusted model, variables with p < 0.20 were included in bivariate analyses, in which only the significant analyses were maintained (p < 0.05). Model adjustment was assessed by residue distribution analysis, which showed normality.

Data were analyzed in Stata, version 14.0, via its survey module, considering our sample design. Factor analysis and linear regression were stratified by age in months and performed for our whole sample.

PNS was approved by the National Research Ethics Commission (Protocol 328,159, of June 26, 2013). All interviewees were contacted and informed of the research and those who agreed to participate in this research signed informed consent forms.

RESULTS

We evaluated 3,646 Brazilian children between six and 23 months (1,249 from six to 11 months; 2,397 from 12 to 23 months). Most were Black or Brown (51.9%) boys (53.3%) aged between 12 and 23 months (68.3%) and belonging to the first and second quintiles of our per capita family income distribution (52.6%). Children living in urban areas (83.8%) in the Brazilian Northeast (29.4%) and Southeast (37.2%) without health insurances (73.6%) and who had visited basic health units (55.4%) and private services (27.9%) showed the highest distribution frequencies (Table 1).

Table 2 describes these children’s food consumption. We observed that the highest percentages of consumed fresh or minimally processed food consisted of cereals and derivatives (83.1%), fruits or natural fruit juices (81.5%), and non-maternal milk and its derivatives (80.5%). The most consumed ultra-processed foods were cookies/cake (76.6%) and more than half of children under two years of age (52.4%) had consumed artificial juices and soft drinks within 24 hours before data collection.

Our data adequacy assessment showed a 0.80 KMO value and Bartlett results with p < 0.001, indicating adequacy for factor analysis.

Our analysis showed two factors, both in our total sample and in the stratified one according to age. Both factors explained 46% of the shared variance in children between six and 11 months and 41% in older ones. Table 3 shows the rotated factor loadings in each of the two components for the total sample and the stratified one. Consumption of “cookies or cake” showed cross-loading. Thus, we removed it from our analysis to better distinguish the validity between factors.

In both age groups and the total sample, the first component – “minimally processed pattern” – included the consumption of fresh or minimally processed foods (non-maternal milk or derivatives, fruits or natural juices, vegetables, beans or other legumes, meat or eggs, potatoes and other tubers and roots, and cereals and derivatives). The second one, “ultra-processed pattern,” included ultra-processed foods (artificial juices, sweets, candies or other sugary foods and soft drinks).

Our crude models showed that adherence to the minimally processed pattern in both age groups was higher among urban white children in higher-income families who resided in the more developed regions of Brazil, had health insurances, and had visited in private health services. Our adjusted models for children aged six to 11 months maintained this pattern, which showed significant associations with higher-income families who resided in urban areas in the more developed Brazilian regions. The model for children aged 12 to 24 months showed that their greater adherence to the minimally processed pattern was significantly associated with white children from higher-income families who resided in more developed regions and visited private health services (Tables 4 and 5).

Regarding ultra-processed patterns among children aged six to 11 months, greater adherence was directly associated with children covered by health insurance and inversely associated with Yellow or Indigenous ones from lower-income families with who lived in the Brazilian Northeast and visited private health services. In children aged one year or older, greater adherence to ultra-processed patterns was directly associated with Black or Brown children residing in the Brazilian South and inversely associated with the richest ones who lived in the Northeast and visited private health services. After adjustment, the associations for children aged six to 11 months remained significant, except for lower incomes and health insurance possession. Children older than one year showed an association which lost significance for lower incomes and private health services. However, we found associations with the Brazilian Southeast and Midwest (Tables 4 and 5).

DISCUSSION

Representative data for Brazilian children under two years of age enabled us to find two food intake patterns, which we classified as minimally processed and ultra-processed. The greater adherence of children between six and 11 months to the minimally processed pattern related to higher family incomes and urban residencies in the more developed regions of Brazil (South, Southeast, and Midwest). In children aged 12 months or older from higher family incomes and living in more developed regions showed higher adherence, an association we also found among those who were white and used private childcare services. Regarding ultra-processed patterns among children aged six to 11 months, greater adherence was inversely associated with Yellow or Indigenous children who lived in the Brazilian Northeast and used private health services. At 12 months or more, greater adherence was directly associated with Black or Brown children who resided in more developed regions, and inversely associated with those living in the Brazilian Northeast.

The minimally processed pattern we found explained the highest proportion of total variance and best represented the food intake of Brazilian children under two years of age. This pattern showed foods or derivatives related to non-breast milk, fruits, meat or eggs, and vegetables in general. Our findings add information to previous findings from some regional studies conducted in Brazil and abroad with children of different ages. These studies also found a type of dietary pattern, consisting mostly of recommended food groups¹⁶16. Bogea EG, Martins MLB, Carvalho WRC, Arruda SPM, França AKTC, Silva AAM. Padrões alimentares de crianças de 13 a 35 meses de idade e associação com características maternas. Cad Saude Publica. 2019;35(4):e00072618. https://doi.org/10.1590/0102-311x00072618
https://doi.org/10.1590/0102-311x0007261... ,²²22. Mais LA, Warkentin S, Vega JB, Latorre MRDO, Carnell S, Taddei JAAC. Sociodemographic, anthropometric and behavioural risk factors for ultra-processed food consumption in a sample of 2–9-year-olds in Brazil. Public Health Nutr. 2018;21(1):77-86. https://doi.org/10.1017/s1368980017002452
https://doi.org/10.1017/s136898001700245... . An example was the “healthy pattern” found in children aged 13 to 35 months in a Northeastern capital, which included vegetables, tubers, meat, offal, rice, pasta, fruit, and fruit juices¹⁶16. Bogea EG, Martins MLB, Carvalho WRC, Arruda SPM, França AKTC, Silva AAM. Padrões alimentares de crianças de 13 a 35 meses de idade e associação com características maternas. Cad Saude Publica. 2019;35(4):e00072618. https://doi.org/10.1590/0102-311x00072618
https://doi.org/10.1590/0102-311x0007261... . A study with two- to nine-year-old schoolchildren in the Brazilian Southeast observed another similar food pattern, “traditional food,” which consisted of six groups of consumed foods: meat, grains, beans, milk and dairy products, vegetables, and fruits²²22. Mais LA, Warkentin S, Vega JB, Latorre MRDO, Carnell S, Taddei JAAC. Sociodemographic, anthropometric and behavioural risk factors for ultra-processed food consumption in a sample of 2–9-year-olds in Brazil. Public Health Nutr. 2018;21(1):77-86. https://doi.org/10.1017/s1368980017002452
https://doi.org/10.1017/s136898001700245... . Moreover, research with Australian children under the age of two²³23. Bell LK, Schammer C, Devenish G, Ha D, Thomson MW, Spencer JA, et al. Dietary patterns and risk of obesity and early childhood caries in Australian toddlers: findings from an Australian cohort study. Nutrients. 2019;11(11):2828. https://doi.org/10.3390/nu11112828
https://doi.org/10.3390/nu11112828... and with European children between two and nine years of age²⁴24. González-Gil EM, Tognon G, Lissner L, Intemann T, Pala V, Galli C, et al. Prospective associations between dietary patterns and high sensitivity C-reactive protein in European children: the IDEFICS study. Eur J Nutr. 2018;57:1397-407. https://doi.org/10.1007/s00394-017-1419-x
https://doi.org/10.1007/s00394-017-1419-... also found patterns consisting of food groups resembling our minimally processed pattern.

Brazilian children tend toward a healthy eating pattern, aligned with the recommendations of the Guia Alimentar para Crianças Brasileiras Menores de 2 anos (Dietary Guidelines for Brazilian Children under 2 Years of Age¹1. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Promoção da Saúde. Guia alimentar para crianças menores de dois anos. Brasília, DF; 2019.). These Guidelines recommend that, from six months of age onward parents or guardians should introduce their children to an adequate and healthy complementary diet to breast milk, based on fresh or minimally processed foods directly obtained from plants and animals, such as fruits, vegetables, legumes, eggs, meats, tubers, grains, and cereals¹1. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Promoção da Saúde. Guia alimentar para crianças menores de dois anos. Brasília, DF; 2019.. Parents and guardians should also start encouraging and adopting healthy eating practices and preventing chronic diseases in children’s later stages of life from their first two years of age onward since eating practices acquired during childhood tend to continue in adulthood²⁵25. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Atenção Básica. Saúde da criança: aleitamento materno e alimentação complementar. 2. ed. Brasília, DF; 2015..

Regarding the consumption of non-breast milk or its derivatives, within the minimally processed pattern, we should mention that parents or guardians should avoid supplying children under one year of age with non-breast milk¹1. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Promoção da Saúde. Guia alimentar para crianças menores de dois anos. Brasília, DF; 2019.. In special situations, a qualified healthcare provider should guide the supply of non-breast milk¹1. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Promoção da Saúde. Guia alimentar para crianças menores de dois anos. Brasília, DF; 2019.. However, as it was impossible to evaluate whether the consumption of this food group followed the recommendations of the Ministry of Health¹1. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Promoção da Saúde. Guia alimentar para crianças menores de dois anos. Brasília, DF; 2019., research should insert the minimally processed dietary pattern in its analyses with caution.

The second pattern we found, “ultra-processed pattern,” included artificial juices, sweets, candies or other sugary foods and soft drinks. Similarly, a study with children aged 13 to 35 months, conducted in Brazilian Northeast urban households, found a dietary pattern its authors called “unhealthy”, which included artificial juices, soft drinks, cookies, simple cakes, and unhealthy snacks¹⁶16. Bogea EG, Martins MLB, Carvalho WRC, Arruda SPM, França AKTC, Silva AAM. Padrões alimentares de crianças de 13 a 35 meses de idade e associação com características maternas. Cad Saude Publica. 2019;35(4):e00072618. https://doi.org/10.1590/0102-311x00072618
https://doi.org/10.1590/0102-311x0007261... . A study conducted with schoolchildren from the Southeast found a pattern its authors called “ultra-processed foods,” consisting of fast food, artificial juice, snacks, sugary snacks, cookies/cakes with filling, and lower vegetable intake²²22. Mais LA, Warkentin S, Vega JB, Latorre MRDO, Carnell S, Taddei JAAC. Sociodemographic, anthropometric and behavioural risk factors for ultra-processed food consumption in a sample of 2–9-year-olds in Brazil. Public Health Nutr. 2018;21(1):77-86. https://doi.org/10.1017/s1368980017002452
https://doi.org/10.1017/s136898001700245... . Note that some differences in the composition of dietary patterns between these studies may result from the different instruments available to obtain data on children’s food intake, the evaluated age group, and the region in which authors conducted their studies, which also increases the relevance of performing this study in a nationally representative sample of children under two years of age.

Since we found a general high prevalence of consumption of many ultra-processed foods, we stress the importance these findings may represent in the health of children under two years of age. Our results indicate that more than three quarters of these children had consumed cookies or cake, data higher than that in a previous study¹⁴14. Dallazen C, Silva SA, Gonçalves VSS, Nilson EAF, Crispim SP, Lang RMF, et al. Introdução de alimentos não recomendados no primeiro ano de vida e fatores associados em crianças de baixo nível socioeconômico. Cad Saude Publica. 2018;34(2):e00202816. https://doi.org/10.1590/0102-311x00202816
https://doi.org/10.1590/0102-311x0020281... . Also, more than half of the children we evaluated consumed sugary drinks, agreeing with results from a study conducted in the Brazilian Southeast²⁶26. Longo-Silva G, Toloni MHA, Menezes RCE, Asakura L, Oliveira MAA, Taddei JAAC. Introdução de refrigerantes e sucos industrializados na dieta de lactentes que frequentam creches públicas. Rev Paul Pediatr. 2015;33(1):34-41. https://doi.org/10.1016/j.rpped.2014.06.009
https://doi.org/10.1016/j.rpped.2014.06.... . Moreover, we observed that about 40% of Brazilian children under two years of age consumed sweets, candies or other sugary foods; a pattern observed in children before four months of age¹⁴14. Dallazen C, Silva SA, Gonçalves VSS, Nilson EAF, Crispim SP, Lang RMF, et al. Introdução de alimentos não recomendados no primeiro ano de vida e fatores associados em crianças de baixo nível socioeconômico. Cad Saude Publica. 2018;34(2):e00202816. https://doi.org/10.1590/0102-311x00202816
https://doi.org/10.1590/0102-311x0020281... . The consumption of ultra-processed foods is associated with unfavorable outcomes to childhood health, such as obesity and dyslipidemias⁷7. Costa CS, Rauber F, Leffa PS, Sangalli CN, Campagnolo PDB, Vitolo MR. Ultra-processed food consumption and its effects on anthropometric and glucose profile: a longitudinal study during childhood. Nutr Metab Cardiovasc Dis. 2019;29(2):177-84. https://doi.org/10.1016/j.numecd.2018.11.003
https://doi.org/10.1016/j.numecd.2018.11... ,⁸8. Rauber F, Campagnolo P, Hoffman D, Vitolo M. Consumption of ultra-processed food products and its effects on children’s lipid profiles: a longitudinal study. Nutr Metab Cardiovasc Dis. 2015;25(1):116-22. https://doi.org/10.1016/j.numecd.2014.08.001
https://doi.org/10.1016/j.numecd.2014.08... ,²⁴24. González-Gil EM, Tognon G, Lissner L, Intemann T, Pala V, Galli C, et al. Prospective associations between dietary patterns and high sensitivity C-reactive protein in European children: the IDEFICS study. Eur J Nutr. 2018;57:1397-407. https://doi.org/10.1007/s00394-017-1419-x
https://doi.org/10.1007/s00394-017-1419-... .

As described in the Guia Alimentar para Crianças Brasileiras Menores de 2 anos (Dietary Guidelines for Brazilian Children under 2 Years of Age¹1. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Promoção da Saúde. Guia alimentar para crianças menores de dois anos. Brasília, DF; 2019.), the ultra-processed dietary pattern we identified is unrecommended for children under two years of age. Although the literature has reported the damage associated with consuming ultra-processed foods⁷7. Costa CS, Rauber F, Leffa PS, Sangalli CN, Campagnolo PDB, Vitolo MR. Ultra-processed food consumption and its effects on anthropometric and glucose profile: a longitudinal study during childhood. Nutr Metab Cardiovasc Dis. 2019;29(2):177-84. https://doi.org/10.1016/j.numecd.2018.11.003
https://doi.org/10.1016/j.numecd.2018.11... ,²⁴24. González-Gil EM, Tognon G, Lissner L, Intemann T, Pala V, Galli C, et al. Prospective associations between dietary patterns and high sensitivity C-reactive protein in European children: the IDEFICS study. Eur J Nutr. 2018;57:1397-407. https://doi.org/10.1007/s00394-017-1419-x
https://doi.org/10.1007/s00394-017-1419-... ,²⁷27. Bigornia SJ, LaValley MP, Noel SE, Moore LL, Ness AR, Newby PK. Sugar-sweetened beverage consumption and central and total adiposity in older children: a prospective study accounting for dietary reporting errors. Public Health Nutr. 2015;18(7):1155-63. https://doi.org/10.1017/s1368980014001700
https://doi.org/10.1017/s136898001400170... , findings in Brazil and abroad indicate the growing consumption of this food group, especially among children in different geographic regions and socioeconomic scenarios²⁶26. Longo-Silva G, Toloni MHA, Menezes RCE, Asakura L, Oliveira MAA, Taddei JAAC. Introdução de refrigerantes e sucos industrializados na dieta de lactentes que frequentam creches públicas. Rev Paul Pediatr. 2015;33(1):34-41. https://doi.org/10.1016/j.rpped.2014.06.009
https://doi.org/10.1016/j.rpped.2014.06.... ,²⁸28. Pan L, Li R, Park S, Galuska DA, Sherry B, Freedman DS. A longitudinal analysis of sugar-sweetened beverage intake in infancy and obesity at 6 years. Pediatrics. 2014;134 Suppl 1:S29-35. https://doi.org/10.1542/peds.2014-0646F
https://doi.org/10.1542/peds.2014-0646F... . The consumption of ultra-processed foods is associated with children’s poor diet, especially those rich in sugar²⁹29. Cornwell B, Villamor E, Mora-Plazas M, Marin C, Monteiro CA, Baylin A. Processed and ultra-processed foods are associated with lower-quality nutrient profiles in children from Colombia. Public Health Nutr. 2018;21(1):142-7. https://doi.org/10.1017/S1368980017000891
https://doi.org/10.1017/S136898001700089... , sodium²⁹29. Cornwell B, Villamor E, Mora-Plazas M, Marin C, Monteiro CA, Baylin A. Processed and ultra-processed foods are associated with lower-quality nutrient profiles in children from Colombia. Public Health Nutr. 2018;21(1):142-7. https://doi.org/10.1017/S1368980017000891
https://doi.org/10.1017/S136898001700089... ,³⁰30. Vandevijvere S, De Ridder K, Fiolet T, Sarah B, Tafforeau J. Consumption of ultra-processed food products and diet quality among children, adolescents and adults in Belgium. Eur J Nutr. 2019;58(8):3267-78. https://doi.org/10.1007/s00394-018-1870-3
https://doi.org/10.1007/s00394-018-1870-... , and saturated fats³⁰30. Vandevijvere S, De Ridder K, Fiolet T, Sarah B, Tafforeau J. Consumption of ultra-processed food products and diet quality among children, adolescents and adults in Belgium. Eur J Nutr. 2019;58(8):3267-78. https://doi.org/10.1007/s00394-018-1870-3
https://doi.org/10.1007/s00394-018-1870-... and those low in vitamins²⁶26. Longo-Silva G, Toloni MHA, Menezes RCE, Asakura L, Oliveira MAA, Taddei JAAC. Introdução de refrigerantes e sucos industrializados na dieta de lactentes que frequentam creches públicas. Rev Paul Pediatr. 2015;33(1):34-41. https://doi.org/10.1016/j.rpped.2014.06.009
https://doi.org/10.1016/j.rpped.2014.06.... ,³⁰30. Vandevijvere S, De Ridder K, Fiolet T, Sarah B, Tafforeau J. Consumption of ultra-processed food products and diet quality among children, adolescents and adults in Belgium. Eur J Nutr. 2019;58(8):3267-78. https://doi.org/10.1007/s00394-018-1870-3
https://doi.org/10.1007/s00394-018-1870-... , fibers, proteins, and potassium⁵5. Rauber F, Costa Louzada ML, Steele EM, Millett C, Monteiro CA, Levy RB. Ultra-processed food consumption and chronic non-communicable diseases-related nutrient profile in the UK (2008-2014). Nutrients. 2018;10(5):587. https://doi.org/10.3390/nu10050587
https://doi.org/10.3390/nu10050587... .

Furthermore, longitudinal studies evince that preschoolers⁷7. Costa CS, Rauber F, Leffa PS, Sangalli CN, Campagnolo PDB, Vitolo MR. Ultra-processed food consumption and its effects on anthropometric and glucose profile: a longitudinal study during childhood. Nutr Metab Cardiovasc Dis. 2019;29(2):177-84. https://doi.org/10.1016/j.numecd.2018.11.003
https://doi.org/10.1016/j.numecd.2018.11... ,⁸8. Rauber F, Campagnolo P, Hoffman D, Vitolo M. Consumption of ultra-processed food products and its effects on children’s lipid profiles: a longitudinal study. Nutr Metab Cardiovasc Dis. 2015;25(1):116-22. https://doi.org/10.1016/j.numecd.2014.08.001
https://doi.org/10.1016/j.numecd.2014.08... ,²⁴24. González-Gil EM, Tognon G, Lissner L, Intemann T, Pala V, Galli C, et al. Prospective associations between dietary patterns and high sensitivity C-reactive protein in European children: the IDEFICS study. Eur J Nutr. 2018;57:1397-407. https://doi.org/10.1007/s00394-017-1419-x
https://doi.org/10.1007/s00394-017-1419-... and adolescents²⁷27. Bigornia SJ, LaValley MP, Noel SE, Moore LL, Ness AR, Newby PK. Sugar-sweetened beverage consumption and central and total adiposity in older children: a prospective study accounting for dietary reporting errors. Public Health Nutr. 2015;18(7):1155-63. https://doi.org/10.1017/s1368980014001700
https://doi.org/10.1017/s136898001400170... in Brazil and other countries consume ultra-processed foods, a diet associated with increased central adiposity⁷7. Costa CS, Rauber F, Leffa PS, Sangalli CN, Campagnolo PDB, Vitolo MR. Ultra-processed food consumption and its effects on anthropometric and glucose profile: a longitudinal study during childhood. Nutr Metab Cardiovasc Dis. 2019;29(2):177-84. https://doi.org/10.1016/j.numecd.2018.11.003
https://doi.org/10.1016/j.numecd.2018.11... ,²⁷27. Bigornia SJ, LaValley MP, Noel SE, Moore LL, Ness AR, Newby PK. Sugar-sweetened beverage consumption and central and total adiposity in older children: a prospective study accounting for dietary reporting errors. Public Health Nutr. 2015;18(7):1155-63. https://doi.org/10.1017/s1368980014001700
https://doi.org/10.1017/s136898001400170... , altered lipid profile⁸8. Rauber F, Campagnolo P, Hoffman D, Vitolo M. Consumption of ultra-processed food products and its effects on children’s lipid profiles: a longitudinal study. Nutr Metab Cardiovasc Dis. 2015;25(1):116-22. https://doi.org/10.1016/j.numecd.2014.08.001
https://doi.org/10.1016/j.numecd.2014.08... , and increased levels of C-reactive protein, a biomarker commonly associated with adiposity and cardiovascular risk factors²⁴24. González-Gil EM, Tognon G, Lissner L, Intemann T, Pala V, Galli C, et al. Prospective associations between dietary patterns and high sensitivity C-reactive protein in European children: the IDEFICS study. Eur J Nutr. 2018;57:1397-407. https://doi.org/10.1007/s00394-017-1419-x
https://doi.org/10.1007/s00394-017-1419-... . The early introduction of ultra-processed foods can also impair exclusive breastfeeding since the supply of this type of food begins before children’s fourth month of life¹⁴14. Dallazen C, Silva SA, Gonçalves VSS, Nilson EAF, Crispim SP, Lang RMF, et al. Introdução de alimentos não recomendados no primeiro ano de vida e fatores associados em crianças de baixo nível socioeconômico. Cad Saude Publica. 2018;34(2):e00202816. https://doi.org/10.1590/0102-311x00202816
https://doi.org/10.1590/0102-311x0020281... . Our results reinforce the need for more effective actions to monitor Brazilian children’s growth and development, especially regarding guidelines aimed at introducing complementary public or private food services.

Analyses performed for both age groups (six to 12 months and 12 to 24 months) found no differences in dietary pattern composition but evinced that Brazilian children already consumed ultra-processed foods as young as one year of age. A result which resembles that from a multicenter study conducted with children living in Southern Brazil¹⁴14. Dallazen C, Silva SA, Gonçalves VSS, Nilson EAF, Crispim SP, Lang RMF, et al. Introdução de alimentos não recomendados no primeiro ano de vida e fatores associados em crianças de baixo nível socioeconômico. Cad Saude Publica. 2018;34(2):e00202816. https://doi.org/10.1590/0102-311x00202816
https://doi.org/10.1590/0102-311x0020281... , which found that the consumption of non-recommended foods begins even before children’s first year of life.

Regarding the adherence of children under two years of age to the minimally processed pattern and higher household incomes, a study conducted in a Brazilian Northeastern capital found similar results¹⁷17. Sotero AM, Cabral PC, Silva GAP. Socioeconomic, cultural and demographic maternal factors associated with dietary patterns of infants. Rev Paul Pediatr. 2015;33(4):445-52. https://doi.org/10.1016/j.rpped.2015.03.006
https://doi.org/10.1016/j.rpped.2015.03.... . However, other studies which used comparable methods failed to find the adherence of white children living in the more developed urban areas of Brazil.

We found that non-white children older than one year showed the highest adherence to the ultra-processed pattern, a result resembling a study with North American children in this age group³¹31. Hamner HC, Perrine CG, Gupta PM, Herrick KA, Cogswell ME. Food consumption patterns among U.S. children from birth to 23 months of age, 2009-2014. Nutrients. 2017;9(9):942. https://doi.org/10.3390/nu9090942
https://doi.org/10.3390/nu9090942... . Regarding the inverse association between the ultra-processed pattern and children living in the Brazilian Northeast, a previous analysis, which evaluated the prevalence of consumption of sugary drinks in children under two years of age via 2013 PNS data³²32. Jaime PC, Prado RR, Malta DC. Influência familiar no consumo de bebidas açucaradas em crianças menores de dois anos. Rev Saude Publica. 2017;51 Supl 1:13s. https://doi.org/10.1590/s1518-8787.2017051000038
https://doi.org/10.1590/s1518-8787.20170... found comparable evidence. However, the ultra-processed dietary pattern not only was inversely associated with children living in the Northeast (as observed in a previous study³²32. Jaime PC, Prado RR, Malta DC. Influência familiar no consumo de bebidas açucaradas em crianças menores de dois anos. Rev Saude Publica. 2017;51 Supl 1:13s. https://doi.org/10.1590/s1518-8787.2017051000038
https://doi.org/10.1590/s1518-8787.20170... ) but was also directly related to children living in more developed regions, suggesting regional differences in the composition of dietary patterns. Unlike a previous finding¹⁴14. Dallazen C, Silva SA, Gonçalves VSS, Nilson EAF, Crispim SP, Lang RMF, et al. Introdução de alimentos não recomendados no primeiro ano de vida e fatores associados em crianças de baixo nível socioeconômico. Cad Saude Publica. 2018;34(2):e00202816. https://doi.org/10.1590/0102-311x00202816
https://doi.org/10.1590/0102-311x0020281... , we observed no association between the ultra-processed pattern and per capita family income.

In general, some health, nutritional, and socioeconomic factors (among others) have made it difficult for children from poor countries to achieve their expected development³³33. Freitas LG, Cortés MAP, Stein C, Cousin E, Faustino-Silva DD, Hilgert JB. Qualidade do consumo alimentar e fatores associados em crianças de um ano de vida na Atenção Primária à Saúde. Cienc Saude Colet. 2020;25(7):2561-70. https://doi.org/10.1590/1413-81232020257.14592018
https://doi.org/10.1590/1413-81232020257... . Research has especially observed that maternal educational attainment and income can interfere in food quality improvement. Moreover, mothers’ educational attainment can influence children’s adherence to both recommended and unrecommended dietary practices³³33. Freitas LG, Cortés MAP, Stein C, Cousin E, Faustino-Silva DD, Hilgert JB. Qualidade do consumo alimentar e fatores associados em crianças de um ano de vida na Atenção Primária à Saúde. Cienc Saude Colet. 2020;25(7):2561-70. https://doi.org/10.1590/1413-81232020257.14592018
https://doi.org/10.1590/1413-81232020257... .

We have so far found no publication evaluating children’s eating patterns and health service use in a nationally representative sample. Our results indicate that the greatest adherence to adequate and healthy eating patterns were associated with children who visited private childcare services. This agrees with a regional study conducted in the Brazilian Northeast¹⁷17. Sotero AM, Cabral PC, Silva GAP. Socioeconomic, cultural and demographic maternal factors associated with dietary patterns of infants. Rev Paul Pediatr. 2015;33(4):445-52. https://doi.org/10.1016/j.rpped.2015.03.006
https://doi.org/10.1016/j.rpped.2015.03.... .

In general, health services, such as childcare, can contribute to several positive outcomes by monitoring children’s growth and development; from promoting and recovering children’s health to preventing diseases in childhood and adulthood¹⁷17. Sotero AM, Cabral PC, Silva GAP. Socioeconomic, cultural and demographic maternal factors associated with dietary patterns of infants. Rev Paul Pediatr. 2015;33(4):445-52. https://doi.org/10.1016/j.rpped.2015.03.006
https://doi.org/10.1016/j.rpped.2015.03.... . Thus, promoting food and nutritional guidance in childcare services is an essential action, especially considering that its lack damages health and is associated with adherence to inadequate eating patterns during childhood⁷7. Costa CS, Rauber F, Leffa PS, Sangalli CN, Campagnolo PDB, Vitolo MR. Ultra-processed food consumption and its effects on anthropometric and glucose profile: a longitudinal study during childhood. Nutr Metab Cardiovasc Dis. 2019;29(2):177-84. https://doi.org/10.1016/j.numecd.2018.11.003
https://doi.org/10.1016/j.numecd.2018.11... ,⁸8. Rauber F, Campagnolo P, Hoffman D, Vitolo M. Consumption of ultra-processed food products and its effects on children’s lipid profiles: a longitudinal study. Nutr Metab Cardiovasc Dis. 2015;25(1):116-22. https://doi.org/10.1016/j.numecd.2014.08.001
https://doi.org/10.1016/j.numecd.2014.08... .

Previous studies indicate that the integrality of childcare is a process under construction in Brazilian Primary Health Care, especially in the services offered by the Family Health Strategy³⁴34. Damasceno SS, Nóbrega VM, Coutinho SED, Reichert APS, Toso BRGO, Collet N. Saúde da criança no Brasil: orientação da rede básica à Atenção Primária à Saúde. Cienc Saude Colet. 2016;21(9):2961-73. https://doi.org/10.1590/1413-81232015219.25002015
https://doi.org/10.1590/1413-81232015219... . Moreover, significant changes in the structure of the service and in professionals’ profile are still needed for an effective Family Health Strategy as a model of care³⁴34. Damasceno SS, Nóbrega VM, Coutinho SED, Reichert APS, Toso BRGO, Collet N. Saúde da criança no Brasil: orientação da rede básica à Atenção Primária à Saúde. Cienc Saude Colet. 2016;21(9):2961-73. https://doi.org/10.1590/1413-81232015219.25002015
https://doi.org/10.1590/1413-81232015219... ,³⁵35. Pedraza DF, Santos IS. Avaliação da vigilância do crescimento nas consultas de puericultura na Estratégia Saúde da Família em dois municípios do estado da Paraíba, Brasil. Epidemiol Serv Saude. 2017;26(4):847-55. https://doi.org/10.5123/s1679-49742017000400015
https://doi.org/10.5123/s1679-4974201700... .

Among the limitations of this study, some were related to the adults taken as a family reference to answer the questionnaire on food consumption. Since they were randomly selected, they may only be indirectly responsible for the child in some households. Furthermore, the use of only one questionnaire on food consumption in the 24 hours prior to data collection may fail to represent children’s usual consumption. However, the use of a large sample with PNS data may have minimized the effect of food consumption variability. Other limitations regard the use of factor analysis, including the following arbitrary options: number of components to be extracted, rotation method, and component naming³⁶36. Martinez ME, Marshall JR, Sochrost L. Invited commentary: factor analysis and the search for objectivity. Am J Epidemiol. 1998;148(1):17-9. https://doi.org/10.1093/oxfordjournals.aje.a009552
https://doi.org/10.1093/oxfordjournals.a... . The observed patterns, however, cohere with the main behaviors which promote chronic non-communicable diseases in Brazil and abroad.

Despite its limitations, this study has important strengths. Our results reinforce the current scientific literature since they found two eating behavior patterns in Brazilian children which are associated to opposite recommendations in the Guia Alimentar para Crianças Brasileiras Menores de 2 anos (Dietary Guidelines for Brazilian Children under 2 Years of Age¹1. Ministério da Saúde (BR), Secretaria de Atenção à Saúde, Departamento de Promoção da Saúde. Guia alimentar para crianças menores de dois anos. Brasília, DF; 2019.). Moreover, we showed that the demographic and economic factors to which children are exposed may influence these patterns. The literature has no national population-based study evaluating outcomes with similar methods (i.e., complex probabilistic sampling and methodological rigor), enabling us to extrapolate our results for the entire infant population under two years of age in Brazil.

Considering our findings, the consumption of some ultra-processed foods showed a high prevalence in Brazilian children under two years of age. We found two dietary patterns of different compositions in this population, which may be distributed according to several sociodemographic characteristics and health service use. Moreover, these patterns seem to represent two opposite recommendations in the Guia Alimentar para Crianças Brasileiras Menores de 2 anos (Dietary Guidelines for Brazilian Children under 2 Years of Age) for food intake and items to be avoided. This may guide the implementation of health promotion actions, including those already in force, so children achieve results with greater equity.

REFERENCES

Publication Dates

History