ABSTRACT

OBJECTIVE

To explore the internal structure and analyze evidence of measurement invariance of the Sisvan form of food intake markers of the Food and Nutrition Surveillance System (Sisvan) for children aged six to 23 months.

METHODS

Sisvan microdata from 2015 to 2019 were used. After analyzing sample adequacy, exploratory factor analysis was carried out considering factor loadings (≥ 0.30 and < 0.85), communality (> 0.20), model fit indices – Tucker-Lewis index (TLI) > 0.90, and root mean square error of approximation (RMSEA) < 0.08. A critical analysis of the form items was performed to adjust the parameters. Measurement stability was investigated between age groups, macro-regions and longitudinally by multi-group confirmatory factor analysis in configural, metric, and scalar invariance models. Configural invariance was accepted if RMSEA < 0.08 and TLI and comparative fit index (CFI) > 0.90. Metric and scalar invariances were accepted if ΔRMSEA < 0.015 and ΔCFI < 0.01, compared to the previous model.

RESULTS

After critical analysis, items were grouped (vegetables and leafy greens; meat or eggs and liver) and excluded (salty food; porridge with milk; rice, potatoes, yams, cassava, flour, or pasta). The exploratory model comprised three factors: healthy complementary feeding (fruit; vegetables; orange or dark green leafy vegetables or fruits; meat, offal, or egg; beans), ultra-processed (yogurt; hamburger and/or sausages; sweetened drinks; instant noodles, packet snacks, or salty cookies; sandwich cookies, sweets, or treats), and milk feeding (breast milk; other milk), with satisfactory factor loadings, communalities and fit (TLI: 0.918, RMSEA: 0.071). The reformulated instrument was stable in the invariance models tested.

CONCLUSIONS

With the incorporation of the modifications to the Sisvan form, the food intake markers subsidize a more adequate interpretation of the feeding situation of children aged six to 23 months.

Food and Nutrition Surveillance; Food Intake; Complementary Feeding; Infants; Validation Study

INTRODUCTION

The first two years of life are crucial for promoting healthy habits¹1. Darling JC, Bamidis PD, Burberry J, Rudolf MC. The first thousand days: early, integrated and evidence-based approaches to improving child health: coming to a population near you? Arch Dis Child. 2020 Sep;105(9):837-41. https://doi.org/10.1136/archdischild-2019-316929
https://doi.org/10.1136/archdischild-201... . Between six and 23 months, the start of complementary feeding concomitant with the continuation of breastfeeding is sensitive to the occurrence of inadequate feeding practices²2. Ministério da Saúde (BR). Secretaria de Atenção Primária à Saúde. Departamento de Promoção da Saúde. Guia alimentar para crianças brasileiras menores de 2 anos. Brasília, DF: Ministério da Saúde; 2019.,³3. UNICEF. The state of the world's children 2019. Children, Food and Nutrition: Growing well in a changing world. New York: Unicef; 2019.. According to the United Nations Children’s Fund, the lack of diversity, low food frequency, and consumption of ultra-processed foods in this period present risks for the development of malnutrition³3. UNICEF. The state of the world's children 2019. Children, Food and Nutrition: Growing well in a changing world. New York: Unicef; 2019..

Demographic and health surveys in 91 low and middle-income countries found that 44.0% of children aged six to 23 months did not eat fruit and vegetables and 42.1% did not eat meat or eggs⁴4. Ricardo LI, Gatica-Domínguez G, Neves PA, Vaz JD, Barros AJ, Wehrmeister FC. Sociodemographic inequalities in vegetables, fruits, and animal source foods consumption in children aged 6-23 months from 91 LMIC. Front Nutr. 2023 Feb;10(1046686):1046686. https://doi.org/10.3389/fnut.2023.1046686
https://doi.org/10.3389/fnut.2023.104668... . Overall, only 27.1% had a diet with the minimum diversity of food groups (95%CI: 23.9-30.3) and 48.7% ate meals at the minimum recommended frequency (95%CI: 45.2-52.2)⁵5. Gatica-Domínguez G, Neves PA, Barros AJ, Victora CG. complementary feeding practices in 80 low- and middle-income countries: prevalence of and socioeconomic inequalities in dietary diversity, meal frequency, and dietary adequacy. J Nutr. 2021 Jul;151(7):1956-64. https://doi.org/10.1093/jn/nxab088
https://doi.org/10.1093/jn/nxab088... .

In the Brazilian National Survey on Child Nutrition (Enani-2019),⁶6. Alves-Santos NH, Castro IR, Anjos LA, Lacerda EM, Normando P, Freitas MB, et al. General methodological aspects in the Brazilian National Survey on Child Nutrition (ENANI-2019): a population-based household survey. Cad Saude Publica. 2021 Aug;37(8):e00300020. https://doi.org/10.1590/0102-311x00300020
https://doi.org/10.1590/0102-311x0030002... ,⁷7. Lacerda EM, Boccolini CS, Alves-Santos NH, Castro IR, Anjos LA, Crispim SP, et al. Methodological aspects of the assessment of dietary intake in the Brazilian National Survey on Child Nutrition (ENANI-2019): a population-based household survey. Cad Saude Publica. 2021 Aug;37(8):e00301420. https://doi.org/10.1590/0102-311x00301420
https://doi.org/10.1590/0102-311x0030142... there was a high prevalence of consumption of ultra-processed foods (80.5%) and exposure to sugar (68.4%) from six to 23 months of age, concomitant with the absence of fruit and vegetable intake in 22.2% of children⁸8. Universidade Federal do Rio de Janeiro. Alimentação infantil I: prevalência de indicadores de alimentação de crianças menores de 5 anos: ENANI 2019. Rio de Janeiro: UFRJ; 2021.. This highlights the need for systematic monitoring of children’s diets.

In Brazil, food and nutrition surveillance (FNS), within the scope of primary health care (PHC) of the Unified Health System (SUS), includes the assessment of food intake markers. For children aged six to 23 months, the Food and Nutrition Surveillance System (Sisvan) focuses on food quality and the risk of nutritional deficiencies and excess weight⁹9. Ministério da Saúde (BR). Secretaria de Atenção à Saúde. Departamento de Atenção Básica. Orientações para avaliação de marcadores de consumo alimentar na atenção básica. Brasília, DF: Ministério da Saúde, 2015 [cited 2023 Sep 7]. Available from: https://bvsms.saude.gov.br/bvs/publicacoes/marcadores_consumo_alimentar_atencao_basica.pdf
https://bvsms.saude.gov.br/bvs/publicaco... .

Although the use of Sisvan food intake markers has been recommended since 2015, there is no evidence of the instrument’s validity, especially regarding its internal structure, in light of dietary recommendations for this age group. In addition, considering how widely the form is used, there is no evidence to support the comparability of information across age groups, given the successive stages of food introduction from six to 23 months, and between the different dietary contexts in the country. The stability of the measurement of the markers over the years, which is relevant to the temporal variations in the population’s dietary practices, is also unknown.

Thus, this study aimed to explore the internal structure of the Sisvan form of food intake markers for children aged six to 23 months and analyze evidence of measurement invariance between age groups, Brazilian macro-regions and longitudinally, from 2015 to 2019.

METHODS

Sisvan Food Intake Markers for Children aged Six to 23 Months and 29 Days

The Sisvan form of food intake markers for children aged six to 23 months and 29 days includes 20 questions referring to the previous day⁹9. Ministério da Saúde (BR). Secretaria de Atenção à Saúde. Departamento de Atenção Básica. Orientações para avaliação de marcadores de consumo alimentar na atenção básica. Brasília, DF: Ministério da Saúde, 2015 [cited 2023 Sep 7]. Available from: https://bvsms.saude.gov.br/bvs/publicacoes/marcadores_consumo_alimentar_atencao_basica.pdf
https://bvsms.saude.gov.br/bvs/publicaco... . For the present study, 17 items on the consumption of foods and/or food groups were considered: “breast milk”, “whole fruit, in pieces or mashed”, “salty food”, “milk other than breast milk”, “porridge with milk”, “yogurt”, “vegetables”, “orange or dark green leafy vegetables or fruits”, “leafy vegetables”, “meat or egg”, “liver”, “beans”, “rice, potatoes, yams, cassava, flour, or pasta”, “hamburgers and/or sausages”, “sweetened drinks”, “instant noodles, packet snacks, or salty cookies”, and “sandwich cookies, sweets, or candies”.

Data Assignment and Management

The General Coordination of Food and Nutrition (CGAN) provided the Sisvan microdata for the period 2015 to 2019, in accordance with Ministry of Health Ordinance No. 884/2011¹⁰10. Ministério da Saúde (BR). Secretaria de Atenção à Saúde. Portaria n° 884, de 13 de dezembro de 2011. Brasília; DF: Ministério da Saúde; 2011 [cited 2023 Sep 7]. Available from: http://bvsms.saude.gov.br/bvs/saudelegis/sas/2011/prt0884_13_12_2011.html
http://bvsms.saude.gov.br/bvs/saudelegis... , and the study was approved by the Research Ethics Committee of the School of Public Health at the University of São Paulo (opinion No. 4.172.787). When managing the data, the records were examined according to: identification number, gender, date of birth, municipality code, and follow-up date, in order to exclude repeated occurrences. In cases of multiple records of the same individual on a single follow-up date (10.35% of observations), the last record of the day was kept. Records of individuals with the same identification number from different municipalities were subtracted (0.44% of observations). Observations with inconsistent age were excluded (0.13% of observations).

For the analyses, only the first record of the individual per year of follow-up was kept, avoiding correlated answers at different points in time. In line with the requirements for psychometric analysis, only records with valid answers (“yes” or “no”) for all items were considered; “don’t know” or blank answers accounted for 9.84% of the records.

The analytical sample comprised 576,034 records of food intake markers for children aged six to 23 months and 29 days between 2015 and 2019. The data was managed using Stata software, version 13.

Analysis of the Internal Structure of the Sisvan Food Intake Markers for Children Aged Six to 23 Months and 29 Days

Evidence of the validity of an instrument falls into five categories: content, response process, internal structure, relationship with other variables, and consequence of testing¹¹11. American Educational Research Association, American Psychologial Association, National Council on Measurement in Education. Standards for educational and psychological testing. Washington, DC: National Council on Measurement in Education; 2014.. This study focused on analyzing the internal structure of food intake markers, using exploratory factor analysis (EFA). In this interdependence technique, variables are grouped into factors that allow latent evaluation dimensions to be identified according to the correlations between them, without previously defining the structure of the data¹²12. Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. Cengage; 2018. Exploratory factor analysis; p. 121-88.,¹³13. Damasio BF. Contributions of the Multigroup confirmatory factor analysis in the invariance evaluation of psychometric tests. Psico-USF. 2013;18(2):211-20. https://doi.org/10.1590/S1413-82712013000200005
https://doi.org/10.1590/S1413-8271201300... . As a prerequisite for the EFA, the sample adequacy analysis was carried out using Bartlett’s sphericity test, which checks the overall statistical significance of the correlations, and the Kaiser-Meyer-Olkin (KMO) test, which indicates the proportion of variance in the data (p < 0.05 and KMO > 0.50, indicative of adequacy, respectively)¹²12. Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. Cengage; 2018. Exploratory factor analysis; p. 121-88..

The number of factors was estimated using parallel analysis. This was followed by EFA with a tetrachoric correlation matrix, promax oblique rotation, and the unweighted least squares extraction method¹²12. Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. Cengage; 2018. Exploratory factor analysis; p. 121-88..

The resulting model was jointly evaluated by factor loadings, communalities and the Tucker-Lewis fit index (TLI), root mean square error of approximation (RMSEA), and Bayesian information criterion (BIC). Factor loadings ≥ 0.30 and < 0.85 and communality > 0.20¹⁴14. Brown TA. Confirmatory factor analysis for applied research. 2nd ed. New York: The Guilford Press; 2015.,¹⁵15. Tabachnick BG, Fidell LS. Using multivariate statistics. 6th ed. Boston: Pearson; 2013. were accepted. In addition, it was recommended that there should be no double saturation or cross-loadings, i.e. items with a significant loading on different factors¹²12. Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. Cengage; 2018. Exploratory factor analysis; p. 121-88.. TLI values > 0.90, RMSEA < 0.08¹⁴14. Brown TA. Confirmatory factor analysis for applied research. 2nd ed. New York: The Guilford Press; 2015.and the lowest BIC value were considered adequate for deciding between the models.

Critical analysis of the form considered compliance with these fit parameters. To this end, some corrective actions were undertaken in this process, followed by re-specification and re-analysis of the model. The following actions were carried out: 1) evaluating variables for exclusion, considering their general contribution to the object analyzed, or for grouping, when there is a similarity between foods and food groups; and 2) changing the number of factors extracted¹²12. Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. Cengage; 2018. Exploratory factor analysis; p. 121-88.. The corrective actions took place sequentially, starting with the critical evaluation of the variables to propose exclusions or groupings. Inadequate factor loadings were then analyzed and, finally, the number of factors extracted was changed. All the EFA stages were carried out using the psych package in the R Studio software, version 4.2.1.

Analysis of Measurement Invariance of Sisvan Food Intake Markers for Children Aged Six to 23 Months and 29 Days

After critically analyzing the form and understanding its internal structure, with satisfactory parameters, an analysis of invariance was carried out in three groups of interest: 1) in age groups (six to < 12 months, 12 to < 18 months, and 18 to < 24 months), to investigate the structure stability in the introduction and continuity of complementary feeding; 2) in the Brazilian macro-regions Midwest, Northeast, North, Southeast, and South, assessing invariance in the light of different feeding practices nationwide; and 3) longitudinally, from 2015 to 2019, looking at the invariance over the years. Multigroup confirmatory factor analysis (MCFA) was used for each section, stipulating the number and distribution of items in the factors,¹²12. Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. Cengage; 2018. Exploratory factor analysis; p. 121-88.,¹³13. Damasio BF. Contributions of the Multigroup confirmatory factor analysis in the invariance evaluation of psychometric tests. Psico-USF. 2013;18(2):211-20. https://doi.org/10.1590/S1413-82712013000200005
https://doi.org/10.1590/S1413-8271201300... ,¹⁶16. Putnick DL, Bornstein MH. Measurement invariance conventions and reporting: the state of the art and future directions for psychological research. Dev Rev. 2016 Sep;41:71-90. https://doi.org/10.1016/j.dr.2016.06.004
https://doi.org/10.1016/j.dr.2016.06.004... in three sequentially restrictive models:

• Model A - Configural invariance or equivalence of form: assesses the plausibility of the number of factors and items per factor;

• Model B - Metric invariance or equivalence of factor loadings: investigates the equivalence of the factor loadings of the items;

• Model C - Scalar invariance or equivalence of intercepts: examines the relationship between the scores obtained and the level of the latent trait.

Configural invariance was accepted when RMSEA < 0.08, TLI and comparative fit index (CFI) > 0.90¹⁴14. Brown TA. Confirmatory factor analysis for applied research. 2nd ed. New York: The Guilford Press; 2015.. Metric and scalar invariance were then tested and accepted when ΔRMSEA < 0.015 and ΔCFI < 0.01 when comparing models B and A and models C and B¹⁶16. Putnick DL, Bornstein MH. Measurement invariance conventions and reporting: the state of the art and future directions for psychological research. Dev Rev. 2016 Sep;41:71-90. https://doi.org/10.1016/j.dr.2016.06.004
https://doi.org/10.1016/j.dr.2016.06.004... . All the MCFA steps were carried out using the lavaan package in the R Studio software, version 4.2.1.

RESULTS

The sample had adequate parameters for the EFA (p < 0.001 in Bartlett’s test of sphericity and KMO = 0.81). The flow of corrective actions in the exploratory phase is shown in Figure 1, with the problems observed summarized in Figure 2.

Considering the 17 items on the form, the first model consisted of five factors and showed adequate TLI and RMSEA values. However, there were inadequate factor loadings and double saturation. A critical analysis of the items was carried out, excluding “salty food” and “porridge with milk”, which were considered non-specific to the object being assessed. It was assumed that the first item could cover a set of subsequent questions on the form, while the second did not make it possible to discriminate the level of processing of the preparation’s ingredients. After this action, model 2 showed adequate TLI and RMSEA values and a lower BIC value when compared to model 1, but with inadequate factor loadings and cross-loadings.

Similar items were grouped together, with “vegetables” and “leafy greens” making up the “vegetables” item, and “meat or egg” and “liver” making up the “meat, offal, or egg” item. Model 3 showed four factors and adequate TLI and RMSEA values, but high factor loadings and double saturation remained.

The next corrective action discarded the item with a high factor loading (“rice, potatoes, yams, cassava, flour, or pasta”). Model 4 showed adequate TLI and RMSEA values, with no cross-loadings and one item with a factor loading above the parameters (“instant noodles, packet snacks, or salty cookies”: 0.87). Finally, this item was excluded, but the model 5 had an inadequate TLI value and a borderline RMSEA value, with inadequate fit.

Model 4 was therefore considered to be the most appropriate solution, accepting the item “instant noodles, packet snacks, or salty cookies” with a slightly higher factor load, recognizing its relevance to characterizing the latent trait of infant feeding practices. The final model had satisfactory fit statistics and consisted of three dimensions (factor 1 - ultra-processed food: yogurt, hamburgers and/or sausages, sweetened drinks, instant noodles, packet snacks, or salty cookies, and sandwich cookies, sweets, or candies; factor 2 - healthy complementary food: whole fruit, vegetables, orange or dark green leafy vegetables or fruits, meat, offal, or egg, and beans; factor 3 - milk feeding: breast milk and milk other than breast milk), as illustrated in Table 1.

Table 2 shows the results of the measurement invariance analyses. The configuration of the model, in terms of number of factors and items per factor, remained equivalent across age groups, macro-regions, and over the years from 2015 to 2019 (model A). Evidence of metric invariance (model B in relation to A) and scalar invariance (model C in relation to B) was also accepted in the three groups of interest, attesting to the stability of the factor loadings of the items and the level of latent trait for the three factors, respectively. The Box illustrates the reformulated proposal for the instrument.

DISCUSSION

This study brings together unprecedented evidence on the internal structure of the Sisvan form of food intake markers for children aged six to 23 months, as well as its invariance for measuring infant feeding practices. The findings confirmed the stability of the factors of healthy complementary feeding, ultra-processed foods, and milk feeding between age groups, Brazilian macro-regions, and longitudinally.

Internationally, there have been efforts to validate instruments for assessing food intake in the first years of life. In Australia, for example, the reliability and validity of a form that classifies the risk of inadequate nutrition in children aged 12 to 36 months, based on food intake in the previous week, has been evaluated¹⁷17. Bell LK, Golley RK, Magarey AM. A short food-group-based dietary questionnaire is reliable and valid for assessing toddlers' dietary risk in relatively advantaged samples. Br J Nutr. 2014 Aug;112(4):627-37. https://doi.org/10.1017/S0007114514001184
https://doi.org/10.1017/S000711451400118... . In the United States, the correspondence between the quality of the diet and the consumption of 14 groups of specific foods and nutrients was analyzed, 10 of which related to the adequacy of the diet (fruits, vegetables, legumes, whole grains, dairy products, animal and vegetable protein sources, seafood, linoleic acid, and alpha-linoleic acid) and four related to the recommendation of moderate consumption (fruit juice, refined grains, sodium, and added sugars)¹⁸18. Kay MC, Duffy EW, Harnack LJ, Anater AS, Hampton JC, Eldridge AL, et al. Development and application of a total diet quality index for toddlers. Nutrients. 2021 Jun;13(6):1943. https://doi.org/10.3390/nu13061943
https://doi.org/10.3390/nu13061943... ,.

In Brazil, Oliveira et al. proposed a form and 18 infant feeding indicators up to two years of age. The form was based on a theoretical model on the attributes, components and markers of complementary feeding¹⁹19. Oliveira JM, Castro IR, Silva GB, Venancio SI, Saldiva SR. Avaliação da alimentação complementar nos dois primeiros anos de vida: proposta de indicadores e de instrumento. Cad Saude Publica. 2015 Feb;31(2):377-94. https://doi.org/10.1590/0102-311X00209513
https://doi.org/10.1590/0102-311X0020951... , to enable the evaluation of indicators recommended by the World Health Organization. However, although efforts were made to adapt the questions to the target audience, there were no assessments of the understanding of the structure, robustness, and validity of the instrument.

The unsatisfactory fit of the original version of the Sisvan form revealed a gap between the instrument’s expectation of objectively assess the phenomenon for which it was proposed and its actual measurement capacity. In the critical analysis that led to corrective actions, the focus was on items that were not specific to the object being assessed and the possibilities of grouping together those with similar characteristics. The internal structure benefited from the reformulation proposal, providing a more rational support for the professional routine of monitoring infant feeding practices in PHC services. The shorter and more precise resulting instrument is in line with the challenges of organizing work in the SUS.

The corrective actions did not affect the estimation of infant feeding indicators. The items maintained with statistical support are satisfactory for constructing the main indicators recommended by international bodies for this age group²⁰20. World Health Organization. Indicators for assessing infant and young child feeding practices: definitions and measurement methods. Geneva: World Health Organization; 2021.. With regard to the exclusion of the item “rice, potatoes, yams, cassava, flour, or pasta (other than instant noodles)”, it should be considered that, as a Brazilian staple food group, the intake of these foods does not particularly differentiate the quality of the diet, which is a characteristic expected of a food intake marker²¹21. England CY, Andrews RC, Jago R, Thompson JL. A systematic review of brief dietary questionnaires suitable for clinical use in the prevention and management of obesity, cardiovascular disease and type 2 diabetes. Eur J Clin Nutr. 2015 Sep;69(9):977-1003. https://doi.org/10.1038/ejcn.2015.6
https://doi.org/10.1038/ejcn.2015.6... . However, as this group is part of international indicators of minimum dietary diversity for children aged six to 23 months²⁰20. World Health Organization. Indicators for assessing infant and young child feeding practices: definitions and measurement methods. Geneva: World Health Organization; 2021., its maintenance in the form is viable (Box) to enable the composition of globally comparable metrics with Sisvan records, although without contributing to the measurement of factors.

After critical analysis, the Sisvan food intake markers covered three crucial factors for dietary recommendations at this stage of life. It is known that the diversity and combination of food groups are important to provide nutritional balance and the evolution of food consistency according to child development²⁰20. World Health Organization. Indicators for assessing infant and young child feeding practices: definitions and measurement methods. Geneva: World Health Organization; 2021.,²²22. Lutter CK, Grummer-Strawn L, Rogers L. Complementary feeding of infants and young children 6 to 23 months of age. Nutr Rev. 2021 Jul;79(8):825-46. https://doi.org/10.1093/nutrit/nuaa143
https://doi.org/10.1093/nutrit/nuaa143... . On the other hand, a diet with excessive amounts of calories, salt, sugar, fats, and additives and especially ultra-processed foods is not recommended up to the age of two. In addition to the nutritional imbalance, these dietary components can reduce children’s interest in fresh foods, de-characterize main meals, and increase the risk of chronic non-communicable diseases²2. Ministério da Saúde (BR). Secretaria de Atenção Primária à Saúde. Departamento de Promoção da Saúde. Guia alimentar para crianças brasileiras menores de 2 anos. Brasília, DF: Ministério da Saúde; 2019.,²⁰20. World Health Organization. Indicators for assessing infant and young child feeding practices: definitions and measurement methods. Geneva: World Health Organization; 2021.,²²22. Lutter CK, Grummer-Strawn L, Rogers L. Complementary feeding of infants and young children 6 to 23 months of age. Nutr Rev. 2021 Jul;79(8):825-46. https://doi.org/10.1093/nutrit/nuaa143
https://doi.org/10.1093/nutrit/nuaa143... . The factors healthy complementary feeding and milk feeding were in line with dietary recommendations for this age group, while the ultra-processed food factor summarized dietary inadequacies in relation to current guidelines.²2. Ministério da Saúde (BR). Secretaria de Atenção Primária à Saúde. Departamento de Promoção da Saúde. Guia alimentar para crianças brasileiras menores de 2 anos. Brasília, DF: Ministério da Saúde; 2019.,²⁰20. World Health Organization. Indicators for assessing infant and young child feeding practices: definitions and measurement methods. Geneva: World Health Organization; 2021.

It is also worth noting that the factors underlying the food intake markers for children aged six to 23 months are linked to recent investigations into the dietary patterns of this population group in Brazil. Using representative data from the National Health Survey (2013), Flores et al. identified three patterns in the dietary intake of children up to 23 months: 1) healthy foods (fruit, vegetables, meat and eggs, potatoes/cassava, cereals, cookies, and beans); 2) milk (breast milk, other milk, and porridge); and 3) unhealthy foods (sweets, soft drinks, and artificial juices)²³23. Flores TR, Neves RG, Wendt A, Costa CD, Bertoldi AD, Nunes BP. Padrões de consumo alimentar em crianças menores de dois anos no Brasil: Pesquisa Nacional de Saúde, 2013. Cien Saude Colet. 2021 Feb;26(2):625-36. https://doi.org/10.1590/1413-81232021262.13152020
https://doi.org/10.1590/1413-81232021262... . Carvalho et al., in turn, observed two patterns for children aged six to 23 months: 1) minimally processed, characterized by the consumption of non-breast milk or dairy products, fruit or natural juice, vegetables, beans or other legumes, meat or eggs, potatoes and other tubers and roots, and cereals and dairy products; and 2) ultra-processed, consisting of the consumption of artificial juices, sweets, candies or other foods with sugar, and soft drinks²⁴24. Carvalho RB, Louzada ML, Rauber F, Levy RB. Characteristics associated with dietary patterns in Brazilian children under two years of age. Rev Saude Publica. 2023 Jan;56(118):118. https://doi.org/10.11606/s1518-8787.2022056003757
https://doi.org/10.11606/s1518-8787.2022... . These findings corroborate that the internal structure of the Sisvan form of food intake markers reflects key points in the diet of children aged six to 23 months.

According to the MCFA, the invariance of items, factor loadings and latent trait scores between age groups, Brazilian macro-regions and longitudinally was confirmed. These findings are supported by factors that denoted a transition in the consistency and composition of the diet, as one factor made up of liquid dairy items and two factors made up of a set of pasty, semi-solid, and solid foods²2. Ministério da Saúde (BR). Secretaria de Atenção Primária à Saúde. Departamento de Promoção da Saúde. Guia alimentar para crianças brasileiras menores de 2 anos. Brasília, DF: Ministério da Saúde; 2019.. It can therefore be assumed that the items in the instrument adequately cover age groups in the successive phases of complementary feeding.

Our finding also suggested that the Sisvan food intake markers bring together questions about infant feeding practices that have capillarity in multiple scenarios in which PHC services are inserted among the Brazilian macro-regions, also capturing dietary characteristics that have remained relevant in recent years. According to the Brazilian Household Budged Survey (POF) 2017-2018, 48.7% of the total calories available in the household came from fresh and minimally processed foods and 19.4% from ultra-processed foods, varying from 11.9% in the North to 23.5% in the South. Compared to data from the 2002-2003 POF, in 2017-2018 there was an increasing trend in the acquisition and the share of ultra-processed foods in total calories consumed (+0.31 percentage points/year), to the detriment of fresh and minimally processed foods (-0.15 percentage points/year)²⁵25. Levy RB, Andrade GC, Cruz GL, Rauber F, Louzada ML, Claro RM, et al. Três décadas da disponibilidade domiciliar de alimentos segundo a NOVA - Brasil, 1987-2018. Rev Saude Publica. 2022;56:75. https://doi.org/10.11606/s1518-8787.2022056004570
https://doi.org/10.11606/s1518-8787.2022... .

Although the POF does not directly assess the food intake of children aged six to 23 months, the relative share of food via household acquisition can influence individual infant feeding practices. According to the 2015 Pelotas (RS) birth cohort, the ultra-processed foods most consumed by children aged six to 23 months were: yogurt (88.3%), boxed juice, powdered juice or coconut water (65.8%), sweet or sandwich cookies (64.5%), candies, lollipops, chocolate, or jelly (64.4%), packet snacks (46.0%), breaded chicken, hamburger, or ultra-processed meats (43.1%), chocolate milk (43.0%), soft drinks (37.4%), and instant noodles (29.6%)²⁶26. Pereira AM, Buffarini R, Domingues MR, Barros FC, Silveira MF. Ultra-processed food consumption by children from a Pelotas Birth Cohort. Rev Saude Publica. 2022 Sep;56(79):79. https://doi.org/10.11606/s1518-8787.2022056003822
https://doi.org/10.11606/s1518-8787.2022... . Similarly, the MINA-Brazil birth cohort, in the northern region of the country, observed frequent consumption of: cookies (66.4%), industrialized yogurt (53.2%), sweets (18.1%), packet snacks (17.9%), and artificial juice (14.7%) at 12 months of age²⁷27. Nogueira MB, Mazzucchetti L, Mosquera PS, Cardoso MA, Malta MB. Consumo de alimentos ultraprocessados e fatores associados no primeiro ano de vida em Cruzeiro do Sul, Acre, Brasil. Cien Saude Colet. 2022 Feb;27(2):725-36. https://doi.org/10.1590/1413-81232022272.47072020
https://doi.org/10.1590/1413-81232022272... .

In children aged six to 23 months, Enani-2019 revealed, for example, a low prevalence of intake of food rich in vitamin A (38.6%), along with a high consumption of sweetened drinks (24.5%), with no significant differences between macro-regions⁸8. Universidade Federal do Rio de Janeiro. Alimentação infantil I: prevalência de indicadores de alimentação de crianças menores de 5 anos: ENANI 2019. Rio de Janeiro: UFRJ; 2021.. This diffuse panorama of dietary inadequacies supports the uniformity of the measurement characteristics of Sisvan food intake markers in the age group investigated between Brazilian macro-regions and longitudinally.

The results of this study should be interpreted considering its limitations. As this was an analysis of a form that had already been implemented nationally for FNS actions, it was possible to explore evidence of validity and invariance of the instrument’s internal structure, without carrying out primary analyses of content validity and the item response process. The record of Sisvan food intake markers has low population coverage²⁸28. Ricci JM, Romito AL, Silva SA, Carioca AA, Lourenço BH. Marcadores do consumo alimentar do Sisvan: tendência temporal da cobertura e integração com o e-SUS APS, 2015-2019. Cien Saude Colet. 2023 Mar;28(3):921-34. https://doi.org/10.1590/1413-81232023283.10552022en
https://doi.org/10.1590/1413-81232023283... but the sample size fully satisfied all requirements for psychometric analyses. Among the strengths are the availability of nationwide data, with responses to the form in situations for which it was designed, from the PHC services of the SUS.

This study joins other efforts to validate the Sisvan food intake markers²⁹29. Lourenço BH, Guedes BM, Santos TS. Marcadores do consumo alimentar do Sisvan: estrutura e invariância de mensuração no Brasil. Rev Saude Publica. 2023;57(1):52. https://doi.org/10.11606/s1518-8787.2023057004896
https://doi.org/10.11606/s1518-8787.2023... . Evidence can contribute to the dissemination of the use of the form in epidemiological research and population surveys among children from six to 23 months, as an instrument with consistent results in terms of its internal structure and measurement equivalence in the groups analyzed. The results of this study shall also be integrated into the expanded concept of FNS, as recommended by the Brazilian National Food and Nutrition Policy³⁰30. Ministério da Saúde (BR). Secretaria de Atenção à Saúde. Departamento de Atenção Básica. Política nacional de alimentação e nutrição. Brasília, DF: Ministério da Saúde, 2013 [cited 2023 Sep 10]. Available from: https://bvsms.saude.gov.br/bvs/publicacoes/politica_nacional_alimentacao_nutricao.pdf
https://bvsms.saude.gov.br/bvs/publicaco... , while strengthening the analysis of data produced continuously by health workers and the systematic food and nutrition monitoring, especially for children under two years of age.

CONCLUSION

After critical analysis and corrective actions, the Sisvan food intake markers for children aged six to 23 months reflected crucial aspects of infant feeding practices at this stage of life. The three complementary feeding factors identified are linked to current recommendations for health promotion from the earliest years and showed stability in configuration, factor loadings, and scores between age groups, Brazilian macro-regions, and longitudinally. The internal structure and evidence of invariance in the reformulated version of the instrument support the use of the markers to monitor food intake during complementary feeding.

REFERENCES

Publication Dates

History