Abstracts

Introduction

Brazil experiences, as other countries, the effects of the double burden of diseases caused by eating errors: nutritional deficiencies in addition to obesity and comorbidities ¹1. Conde WL, Monteiro CA. Nutrition transition and double burden of undernutrition and excess of weight in Brazil. Am J Clin Nutr 2014; 100:1617S-22S.^,22. Rivera JA, Pedraza LS, Martorell R, Gil A. Introduction to the double burden of undernutrition and excess weight in Latin America. Am J Clin Nutr 2014; 100:1613S-6S.^,33. World Cancer Research Fund; American Institute for Cancer Research. Diet, nutrition, physical activity and cancer: a global perspective. Continuous update project expert report. Washington DC: American Institute for Cancer Research; 2018.; while 2.7% of the infant population shows low weight, obesity already reaches 14.8% of the population ⁴4. Ministério da Saúde. Pesquisa de Orçamentos Familiares 2008-2009. Avaliação nutricional da disponibilidade de alimentos no Brasil. Brasília: Ministério da Saúde; 2010.. Several studies bring evidence of relationships between consumption of specific food groups, as well as their degree of processing and the occurrence of chronic non-communicable diseases. These findings subsidize the national and international dietary guidelines for the prevention of chronic non-communicable diseases ⁵5. GBD 2017 Diet Collaborators. Health effects of dietary risks in 195 countries, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2019; 393:1958-72..

Food guides provide guidelines to assist citizens, the productive sector, health professionals, and public administrators in promoting adequate nutrition ⁶6. Departamento de Atenção Básica, Secretaria de Atenção à Saúde, Ministério da Saúde. Guia alimentar para a população brasileira. Brasília: Ministério da Saúde; 2014.. However, adjusting eating habits requires a multi-factor long-term approach, and monitoring and evaluating dietary intake may promote public policies that favor progress towards healthy choices ⁵5. GBD 2017 Diet Collaborators. Health effects of dietary risks in 195 countries, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2019; 393:1958-72..

Since 1980, indexes or other types of measures have been developed and their association with the guides has been tried ⁷7. Dixon LB. Updating the healthy eating index to reflect current dietary guidance. J Am Diet Assoc 2008; 108:1837-42.. Diet quality indexes were developed for food intake monitoring and are also used as a tool for food and nutrition education ⁸8. Kennedy ET, Ohls J, Carlson S, Fleming K. The healthy eating index: design and applications. J Am Diet Assoc 1995; 95:1103-8.^,99. Guenther PM, Casavale KO, Reedy J, Kirkpatrick S, Hiza HAB, Kuczynski KJ, et al. Update of the healthy eating index: HEI-2010. J Acad Nutr Diet 2013; 113:569-80.^,1010. Guenther PM, Kirkpatrick S, Reedy J, Krebs-Smith SM, Buckman DW, Casavale KO, et al. The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans. J Nutr 2014; 144:399-407.. The Health Eating Index (HEI) ⁹9. Guenther PM, Casavale KO, Reedy J, Kirkpatrick S, Hiza HAB, Kuczynski KJ, et al. Update of the healthy eating index: HEI-2010. J Acad Nutr Diet 2013; 113:569-80. is updated at every revision of the Dietary Guidelines for Americans¹¹11. U.S. Department of Agriculture; U.S. Department of Health and Human Services. Dietary guidelines for Americans 2010. 7th Ed. Washington DC: U.S. Government Printing Office; 2010.. Thus, it incorporates advances regarding the relationship between diet and health and enhances its psychometric properties. Its second version ¹²12. Guenther PM, Reedy J, Krebs-Smith SM. Development and evaluation of the Healthy Eating Index-2005: technical report. Washington DC: Center for Nutrition Policy and Promotion; 2007. was adapted for the use in Brazil and considered the first version of the Food Guide for the Brazilian Population as a reference for its calculations ¹³13. Previdelli AN, Ferreira SRG, Fisberg RM, Marchioni DM. Índice de Qualidade da Dieta revisado para população brasileira. Rev Saúde Pública 2011; 45:794-8..

The Digital Food Guide (DFG) was developed in the years of 2011-2012 as a technological tool that could evaluate the quality of the diet. The creation of the Diet Quality Index (DQI) associated with DFG (DQI-DFG) is the result of studies on nutritional consensus about patterns of food intake compatible with health promotion. Since its launch in 2009, it has contributed to promoting choices in key food groups for good nutrition. In the construction of this index, foods are classified into food groups according to their similarity in nutritional composition, as well as the evidences about the implications of their regular intake on health ¹⁴14. Caivano S, Domene SMA. Diet quality index for healthy food choices. Rev Nutr 2013; 26:693-9.. Based on the criteria established in the HEI-2005 ¹²12. Guenther PM, Reedy J, Krebs-Smith SM. Development and evaluation of the Healthy Eating Index-2005: technical report. Washington DC: Center for Nutrition Policy and Promotion; 2007., in addition to the guidelines from the Department of Nutrition of the Harvard School of Public Health (United States) ¹⁵15. Harvard School of Public Health. Food pyramid and plates: what should you really eat? http://www.hsph.harvard.edu/nutritionsource/pyramid-full-story/ (accessed on Aug/2012).
http://www.hsph.harvard.edu/nutritionsou... ^,1616. Harvard School of Public Health. Harvard researchers launch Healthy Eating Plate. https://www.hsph.harvard.edu/news/press-releases/healthy-eating-plate/ (accessed on Aug/2012).
https://www.hsph.harvard.edu/news/press-... ^,1717. Willett W. Eat, drink, and be healthy: the Harvard Medical School guide to healthy eating. New York: Simon and Schuster; 2011., adjustments were carried out for the valuation of the Brazilian population food habits, and thus, the DQI-DFG was developed (DQI-DFG-2013) ¹⁴14. Caivano S, Domene SMA. Diet quality index for healthy food choices. Rev Nutr 2013; 26:693-9., instead of the indicators that derive from the guidelines of the North American Food Guide.

New findings on the effect of food on health motivate the revision of the criteria for the organization of the moderation components and the adequacy of DQI-DFG. The present study aims to improve and validate the (DQI-DFG-2019).

Methods

To improve and validate the DQI-DFG-2019, its psychometric properties for content validation, construct validation, and reliability assessment were studied this way: (i) Content validation - Consensus from an expert panel about healthy eating and DQI; (ii) Construct validation - Reference diet improvement for determining the portions and ranges of intake of each food group; Correlation between the components of the DQI-DFG and the energy value (Pearson’s linear correlation coefficient); Analysis of agreement between the experts' opinion about the quality of menus and the result of the DQI-DFG (frequency, Pearson’s linear correlation coefficient); (iii) Reliability - Reliability based on the evaluators (intraclass correlation coefficient, Student’s t-test, signal test, and Model of analysis of variance in blocks); Correlation between the components of the DQI-DFG and the total score (Pearson’s linear correlation coefficient).

Database construction

The DQI-DFG-2019 was applied to menus originated from secondary data of the cross-sectional population-based survey named Health Survey in Campinas - IS-ACamp 2008. Food intake was estimated by 24-hour recalls (24hR), collected on different days of the week and months of the year, by a team of trained and supervised interviewers. The criteria for sampling and randomizing the study have been published ¹⁸18. Alves MCGP. Plano de amostragem. In: Barros MBA, César CLG, Carandina L, Goldbaum M, organizadores. As dimensões da saúde: inquérito populacional em Campinas. São Paulo: Editora Hucitec; 2008. p. 46-55.. The sample of this study consisted of 664 adults, aged 19-60 years, of both sexes, who do not practice professional physical exercise, are not in pregnancy and lactation period, and with no presence of disease that demands nutritional intervention. This sample comes from Brazilian individuals, from the southeast region, living in the city of Campinas, State of São Paulo. Data from the 24hR were processed by Nutrition Data System for Research software, 2007 version (http://www.ncc.umn.edu/products/) and exported to the Microsoft Access software (https://products.office.com/) for the calculation of the DQI-DFG-2019.

Content validation

Consensus from an expert panel about healthy eating and DQI

Likert is a scale for perception assessment by indicating agreement or disagreement about what is being measured ¹⁹19. Rowe G, Wright G. The Delphi technique as a forecasting tool: issues and analysis. Int J Forecast 1999; 15:353-75.. Thus, assertions related to the topic were developed and the experts attributed agreement or disagreement. The Delphi technique is adopted to establish consensus among experts ²⁰20. Wakita T, Ueshima N, Noguchi H. Psychological distance between categories in the Likert Scale: comparing different numbers of options. Educ Psychol Meas 2012; 72:533-46.. The frequency of agreement or disagreement of each assertion was presented to the experts in order to confirm the consensus on the subject. In this way, the Likert scale ¹⁹19. Rowe G, Wright G. The Delphi technique as a forecasting tool: issues and analysis. Int J Forecast 1999; 15:353-75. and the Delphi technique ²⁰20. Wakita T, Ueshima N, Noguchi H. Psychological distance between categories in the Likert Scale: comparing different numbers of options. Educ Psychol Meas 2012; 72:533-46. were combined to perform index content validation.

Assertions contained in the analysis instrument followed the same theoretical background used for the construction of the DQI-DFG-2013 ¹⁴14. Caivano S, Domene SMA. Diet quality index for healthy food choices. Rev Nutr 2013; 26:693-9.. These assertions and their respective statistical analyses are contained in the article Consensus Between Specialists in Healthy Eating and Diet Quality Index²¹21. Caivano S, Domene SMA. Consensus among experts on healthy eating and diet quality index. Ciênc Saúde Colet 2018. http://www.cienciaesaudecoletiva.com.br/artigos/consensus-among-experts-on-healthy-eating-and-diet-quality-index/16995?id=16995.
http://www.cienciaesaudecoletiva.com.br/... .

Construct validation

Reference diet for determining the portion sizes and ranges of intake of each food group

The reference diet (RD) developed in the DQI-DFG-201314 was enhanced through the attribution of criteria to define the portion size for each of the food groups and the corresponding intake range. The RD was calculated using the Brazilian Table of Food Composition (TACO) ²²22. Núcleo de Estudos e Pesquisas em Alimentação. TACO - Tabela Brasileira de Composição de Alimentos. 4a Ed. Campinas: Universidade Estadual de Campinas; 2011. and the Database of Nutrients from the U.S. Department of Agriculture (USDA) ²³23. U.S. Department of Agriculture. National Nutrient Database for Standard Reference, release 26. https://ndb.nal.usda.gov/ndb/search/list (accessed on Jun/2011).
https://ndb.nal.usda.gov/ndb/search/list... . The nutritional quotas were determined through the reference values for nutritional requirements proposed by the Institute of Medicine (United States) ²⁴24. Institute of Medicine. Dietary reference intakes for thiamin, riboflavin, niacin, vitamin B6, folate, vitamin B12, pantothenic acid, biotin, and choline. Washington DC: The National Academies Press; 2000.^,2525. Institute of Medicine. Dietary reference intakes for vitamin C, vitamin E, selenium, and carotenoids. Washington DC: The National Academies Press; 2000.^,2626. Institute of Medicine. Dietary reference intakes for vitamin A, vitamin K, arsenic, boron, chromium, copper, iodine, iron, manganese, molybdenum, nickel, silicon, vanadium, and zinc. Washington DC: The National Academies Press; 2001.^,2727. Institute of Medicine. Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids (macronutrients). Washington DC: The National Academies Press; 2002.^,2828. Institute of Medicine. Dietary reference intakes for water, potassium, sodium, chloride, and sulfate. Washington DC: The National Academies Press; 2005.^,2929. Institute of Medicine. Dietary reference intakes for calcium and vitamin D. Washington DC: The National Academies Press; 2010. for macronutrients, vitamins and minerals, linoleic acid, and linolenic acid; for saturated, polyunsaturated, and monounsaturated fats, we employed the guidelines from the Arquivos Brasileiros de Cardiologia³⁰30. Sociedade Brasileira de Cardiologia. IV diretriz brasileira sobre dislipidemias e prevenção da aterosclerose. Arq Bras Cardiol 2007; 88:2-19.. This topic includes the application of nutritional concepts derived from a dietary technique, which will be confirmed through the statistical analysis presented in items Correlation between the Components of the DQI-DFG and the Energy Value and Analysis of Agreement between the Experts’ Opinion about the Quality of Menus and the Result of the DQI-DFG from the construct validation.

Correlation between the components of the DQI-DFG and the energy value

This analysis was applied to ensure the independence between the score in the index components and the energy value of the diet and to avoid the overestimation of the final score in high energy diets.

Analysis of agreement between the experts’ opinion about the quality of menus and the result of the DQI-DFG

The DQI-DFG is organized into moderation components and adequacy components. Score and weight criteria were assigned according to the ranges of portions intake, according to the relevance of each food group. Blind evaluation by experts in HEI and DQI was submitted to analysis of agreement with the final result generated by the DQI-DFG-2019.

a) Selection of experts

A search in the Web of Science database, with keywords “healthy eating index” and “diet quality index”, added with Brazil as the address, was used to identify authors of articles published between June 2005 and June 2015. Through the Lattes platform, provided by the Brazilian National Research Council (CNPq), we selected graduated dietitian experts.

b) Data collection with the experts

The experts received an email with the instructions and the link to access the survey instrument, hosted in the tool Question Pro (https://www.questionpro.com/pt-br/).

c) Analysis instrument

A random selection of ten different menus representatives of each tertile of the DQI-DFG score was sent to experts. In blind analysis, the experts were asked to consider: occurrence of different food groups, concentrated sources of sugar and fat, derived products with high degree of processing and characteristics of preparation associated with risk or protection to classify each menu as “good” “regular”, or “bad”, assigning grades from 0 (very bad) to 10 (very good). The foods and recipes contained in the menus were accompanied by household measure with the objective of assisting in qualitative analysis, without requiring nutritional calculation.

Reliability

Reliability based on the evaluators

In order to assess the reproducibility of the DQI-DFG-2019, three undergraduate students and five nutritionists from the research group received a list of ten different menus chosen randomly. The result of each evaluator for the processed menus was compared with the result of the DQI-DFG-2019 (inter-rater evaluation).

Nutrabem Pro

The Nutrabem Pro (https://nutrabem.unifesp.br/) was the software used to calculate the menus. The Nutrabem Pro software database was built with data from the TACO ²²22. Núcleo de Estudos e Pesquisas em Alimentação. TACO - Tabela Brasileira de Composição de Alimentos. 4a Ed. Campinas: Universidade Estadual de Campinas; 2011. and from the USDA Food Composition Database ²³23. U.S. Department of Agriculture. National Nutrient Database for Standard Reference, release 26. https://ndb.nal.usda.gov/ndb/search/list (accessed on Jun/2011).
https://ndb.nal.usda.gov/ndb/search/list... ; this tool is free of charge and operates the DQI-DFG-2019 application.

Correlation between the components of the DQI-DFG and the total score

This analysis was employed to determine which component exerts greater influence on the score of the Index.

Ethical aspects

This study was approved by the Ethics Research Committee of the São Paulo Federal University - opinion n. 707.010, and by the system CEP/CONEP - CAAE n. 28924514.6.0000.5505.

Results and discussion

The theoretical framework and the conception that support the DQI-DFG were examined in the validation and reliability study, as shown below.

Content validation

Consensus on healthy eating and diet quality index between experts

The result of the consensus between the experts ²¹21. Caivano S, Domene SMA. Consensus among experts on healthy eating and diet quality index. Ciênc Saúde Colet 2018. http://www.cienciaesaudecoletiva.com.br/artigos/consensus-among-experts-on-healthy-eating-and-diet-quality-index/16995?id=16995.
http://www.cienciaesaudecoletiva.com.br/... ratifies the division of the food groups into moderation components and adequacy components (Table 1). Foods classified among the moderation components bring risk to health when consumed in excess. However, they can integrate the menu without compromising the quality of nutrition for possible satisfaction of cultural or sensory values, as ultra-processed foods. The groups that comprise the adequacy components are those with important foods to meet the nutritional needs ¹⁴14. Caivano S, Domene SMA. Diet quality index for healthy food choices. Rev Nutr 2013; 26:693-9.. It is the opinion of the experts that composing a DQI using food groups enables to emphasize foods and not nutrients ²¹21. Caivano S, Domene SMA. Consensus among experts on healthy eating and diet quality index. Ciênc Saúde Colet 2018. http://www.cienciaesaudecoletiva.com.br/artigos/consensus-among-experts-on-healthy-eating-and-diet-quality-index/16995?id=16995.
http://www.cienciaesaudecoletiva.com.br/... , similarly to the precepts of the Food Guide for the Brazilian Population⁶6. Departamento de Atenção Básica, Secretaria de Atenção à Saúde, Ministério da Saúde. Guia alimentar para a população brasileira. Brasília: Ministério da Saúde; 2014.:

(i) Sugars and sweets: sugar-rich foods like sodas, chocolates, ice cream, or cookies can bring risks to health when there is excessive intake; therefore, to be part of a healthy eating, they should not be of frequent intake, and in small portion sizes ³¹31. Martins APB, Levy RB, Claro RM, Moubarac JC, Monteiro CA. Increased contribution of ultra-processed food products in the Brazilian diet (1987-2009). Rev Saúde Pública 2013; 47:656-65.^,3232. Parylak SL, Koob GF, Zorrilla EP. The dark side of food addiction. Physiol Behav 2011; 104:149-56.. Studies on sugar dependence and sweet taste dependence show that excessive intake of sugary foods or only sweet tasting, such as diet and light products, stimulates the pleasure derived from food and the amount of food intake ³³33. Lustig RH, Schmidt LA, Brindis CD. The toxic truth about sugar. Health Policy 2012; 482:7-9.. Therefore, diet and light foods also compose this group.

(ii) Meats - beef, pork, and processed meat: they are important sources of protein and vitamin B12, but their intake is associated with the risk of developing colorectal cancer ³⁴34. American Institute for Cancer Research. Food, nutrition, physical activity and prevention of cancer: a global perspective. Washington DC: World Cancer Research Fund/American Institute for Cancer Research; 2007.; therefore, considering the risk caused by excess intake of these foods, quotas for ingestion must be established.

(iii) Refined cereals: considering their higher glycemic index and association with obesity, type 2 diabetes, metabolic syndrome, and cardiovascular diseases, foods from these groups should have their intake reduced ³⁵35. Katz DL, Meller S. Can we say what diet is best for health? Annu Rev Public Health 2014; 35:83-103.; for the same reason, potatoes are classified into this group ³⁶36. Bach-Faig A, Berry EM, Lairon D, Requant J, Trichopoulou A, Dernini S, et al. Mediterranean diet pyramid today. Science and cultural updates. Public Health Nutr 2011; 14:2274-84.^,3737. Muraki I, Rimm EB, Willett WC, Manson JE, Hu FB, Sun Q. Potato consumption and risk of type 2 diabetes: results from three prospective cohort studies. Diabetes Care 2016; 39:376-84..

(iv) Processed fats: foods with high concentration of saturated fat and sodium, when consumed in excess are associated with the development of cardiovascular diseases, especially on diets high in refined carbohydrates ⁶6. Departamento de Atenção Básica, Secretaria de Atenção à Saúde, Ministério da Saúde. Guia alimentar para a população brasileira. Brasília: Ministério da Saúde; 2014.. For this reason, they are among the moderation components.

(v) Poultry, fish, and eggs: the classification into a distinct group in relation to other meats relies in the understanding of healthy eating, since the intake of red meat and processed meat is associated with the risk of developing colorectal cancer ³⁴34. American Institute for Cancer Research. Food, nutrition, physical activity and prevention of cancer: a global perspective. Washington DC: World Cancer Research Fund/American Institute for Cancer Research; 2007..

(vi) Whole cereals, tubers and roots: they are sources of dietary fiber, bioactive compounds, and have low glycemic index ³⁵35. Katz DL, Meller S. Can we say what diet is best for health? Annu Rev Public Health 2014; 35:83-103..

(vii) Fruits and vegetables: intake of these fresh foods, sources of micronutrients and bioactive compounds, is associated to lower risk for developing cardiovascular diseases and to the prevention of the most chronic diseases ⁶6. Departamento de Atenção Básica, Secretaria de Atenção à Saúde, Ministério da Saúde. Guia alimentar para a população brasileira. Brasília: Ministério da Saúde; 2014.^,3838. Bonaccio M, Iacoviello L, de Gaetano G. The Mediterranean diet: the reasons for a success. Thromb Res 2012; 129:401-4..

(viii) Legumes and oilseeds: beans are part of the Brazilian staple foods ⁶6. Departamento de Atenção Básica, Secretaria de Atenção à Saúde, Ministério da Saúde. Guia alimentar para a população brasileira. Brasília: Ministério da Saúde; 2014. and are sources of proteins, fibers, vitamins, monounsaturated fatty acids, and minerals ¹⁷17. Willett W. Eat, drink, and be healthy: the Harvard Medical School guide to healthy eating. New York: Simon and Schuster; 2011.. Although oilseeds also provide these nutrients, there is no indication that supports the need for their daily intake; they should be part of a healthy diet in order to vary the food repertoire.

(ix) Milk and dairy products: dairy products present high concentrations of nutrients, especially calcium, which justifies its inclusion in the repertoire of healthy food choices ⁶6. Departamento de Atenção Básica, Secretaria de Atenção à Saúde, Ministério da Saúde. Guia alimentar para a população brasileira. Brasília: Ministério da Saúde; 2014..

(x) Oils and fats: they provide essential fatty acids for biochemical processes related to various systems, in addition to being a protection factor for cardiovascular diseases ¹⁷17. Willett W. Eat, drink, and be healthy: the Harvard Medical School guide to healthy eating. New York: Simon and Schuster; 2011.^,3838. Bonaccio M, Iacoviello L, de Gaetano G. The Mediterranean diet: the reasons for a success. Thromb Res 2012; 129:401-4..

(xi) Alcoholic beverage: its consumption is not evaluated by the DQI-DFG-2019. Alcohol is the most used drug in the world and is responsible for 3.2% of deaths in Brazil - a social and public health problem ³⁹39. Laranjeira R, Pinsky I, Zaleski M, Caetano R. I levantamento nacional sobre os padrões de consumo de álcool na população brasileira. Brasília: Secretaria Nacional Antidrogas; 2007.. Although moderate intake of wine is recognized as part of the strategies for the prevention of cardiovascular diseases, alcohol is among the risk factors for cancer ⁴⁰40. Brenner H, Kloor M, Pox CP. Colorectal cancer. Lancet 2014; 383:1490-502..

Construct validation

Improvement of the reference diet for determining portion sizes and ranges of intake of each food group

We developed an RD (2,000Kcal) consisting of typical foods that preserve aspects of culture and preferences of the Brazilian population. The dietary calculation of this RD met the quotas of nutrients and energy density so they are consistent with the reduction of chronic diseases and of the risk of nutritional deficiencies:

(i) Breakfast: 0.5 small unit (151g) of papaya, 8 leveled tablespoons (30g) of skimmed cow’s milk powder, 2 slices (57g) of whole wheat sliced bread, and 4 leveled tablespoons (30g) of creamy curd cheese;

(ii) Mid-morning snack: 1 large unit (110g) of yellow cavendish banana with 1 tablespoon shallow (7g) of flaked oats;

(iii) Lunch: 2 saucers full (50g) of lettuce, 1 medium unit (80g) of salad tomato, 1 teaspoon (2g) of flaxseed, 2 tablespoons (8mL) of extra virgin olive oil, 80g of raw brown rice, 40g of beans, 1 small unit (100g) of raw chicken breast, 0.5 small unit (30g) of raw carrot, 1 medium unit (110g) of orange. Seasonings: 2 teaspoons (2g) of refined salt, 1 medium unit (70g) of onion, 10 cloves (20g) of garlic, 4 teaspoons (8mL) of soy oil.

(iv) Afternoon snack: 1 unit (160g) skimmed natural yogurt with 70g of Palmer mango and 2 dessert spoons (16g) of traditional granola;

(v) Dinner: soup: 2 tablespoons (30g) of chickpeas, 2 saucers (40g) of kale, 2 medium pieces (50g) of pumpkin, 2 medium pieces (130g) of yams, 1 unit (50g) of chicken egg, 1.5 teaspoon (1.5g) of refined salt, 2 teaspoons (3mL) of extra virgin olive oil, 2 teaspoons (3mL) of soy oil and 1 small slice (100mL) of watermelon;

(vi) Night snack: 11 units (3-4g) of Brazilian nuts, 0.5 tablespoon (3-5g) of salted roasted peanuts, and 2 tea cups (180mL) of fennel tea.

The portion sizes for each food group represented the core value for the definition of the adequate intake ranges. Table 2 shows the corresponding energy according to the food groups and the proposed intake ranges.

a) Moderation components

In the DQI-DFG-2013 construct, the score attributed to these components was dichotomous (within the intake range they receive the maximum points - 5 or 10; when the range was exceeded, the score was set to zero) ¹³13. Previdelli AN, Ferreira SRG, Fisberg RM, Marchioni DM. Índice de Qualidade da Dieta revisado para população brasileira. Rev Saúde Pública 2011; 45:794-8.. In this improvement, when the portion is within the intake range, the score is fixed at 0.5 points; if the portion is up to twice the upper limit of the intake range the score will be proportionally decreasing; above twice the value these food groups will not receive points. Thus, we protect the diet from excessive refined carbohydrates, high glycemic index foods, and saturated fats ¹⁵15. Harvard School of Public Health. Food pyramid and plates: what should you really eat? http://www.hsph.harvard.edu/nutritionsource/pyramid-full-story/ (accessed on Aug/2012).
http://www.hsph.harvard.edu/nutritionsou... ^,1616. Harvard School of Public Health. Harvard researchers launch Healthy Eating Plate. https://www.hsph.harvard.edu/news/press-releases/healthy-eating-plate/ (accessed on Aug/2012).
https://www.hsph.harvard.edu/news/press-... .

Twice the upper limit of the intake range: if the range is 0.0-0.5 portions, twice the upper limit is 1.0; to 0.4 portions of sugars and sweets is attributed 0.5 point; if 0.8 portions, proportional descending score (0.2 points); if 1.4 portions, 0 points.

b) Adequacy components

In the DQI-DFG-2013 construct, the score was proportionally increasing up to reaching the minimum value of the intake range; within or above the intake range, each food group received its maximum score (5 or 10 points) ¹⁴14. Caivano S, Domene SMA. Diet quality index for healthy food choices. Rev Nutr 2013; 26:693-9.. In this improvement, portions below and within the intake range remain as in the DQI-DFG-2013. However, if the portion is above the intake range, the score will be different according to the food group: for Whole cereals, tubers, and roots, Fruits, and Milk and dairy products, if the portion is up to twice the upper limit of the intake range the score will be proportionally decreasing; and above twice the upper limit of the intake range, these food groups will not receive points. For the groups Poultry, fish, and eggs, Vegetables, Legumes and oilseeds, and Oils and fats, even with portion values above the intake range, the score remains fixed.

Twice the upper limit of the intake range: if the range is 1.5-3.0 portions, twice the upper limit is 6.0. To 1.3 portions of Fruit is assigned a proportionally increasing score (1.3 points); if 2.0 portions, 1.5 points; if 5.0 portions, proportionally decreasing score (0.5 points); if 6.4 portions, 0 points.

Consuming fruits, legumes, vegetables, whole cereals, poultry, fish, eggs, milk and dairy products, vegetable oils and oilseeds regularly is compatible with a healthy diet, as it ensures the supply of nutrients in order to meet the nutritional needs for adults ¹⁴14. Caivano S, Domene SMA. Diet quality index for healthy food choices. Rev Nutr 2013; 26:693-9.. However, high intake of Whole cereals, tubers, and roots and Fruits can be associated to the development of obesity, type 2 diabetes, metabolic syndrome, and cardiovascular diseases, due to increasing the glycemic index of the whole diet ⁴¹41. Santos FL, Esteves SS, da Costa Pereira A, Yancy Jr. WS, Nunes JP. Systematic review and meta-analysis of clinical trials of the effects of low carbohydrate diets on cardiovascular risk factors. Obes Rev 2012; 13:1048-66.. High intake of Milk and dairy products can exceed the recommendation of 10% for saturated fats ²⁷27. Institute of Medicine. Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids (macronutrients). Washington DC: The National Academies Press; 2002., favoring the development of cardiovascular diseases ⁴²42. Siri-Tarino P, Sun Q, Hu FB, Krauss RM. Saturated fat, carbohydrate, and cardiovascular disease. Am J Clin Nutr 2010; 91:502-9..

DQI generates values between 0 and 100 points: < 40 points, low quality diet; 40-70 points, intermediate quality diet; > 70 points, good quality diet. These cut-off points were defined according to the qualitative characteristic of the study’s menus and can be adjusted according to the criteria defined by other researchers to be adapted to the different patterns of consumption.

Correlation between the components of the DQI-DFG-2019 and the energy value

The organization of the food groups into moderation or adequacy components is adopted by the HEI-2005 ¹²12. Guenther PM, Reedy J, Krebs-Smith SM. Development and evaluation of the Healthy Eating Index-2005: technical report. Washington DC: Center for Nutrition Policy and Promotion; 2007. and HEI-2010 ⁹9. Guenther PM, Casavale KO, Reedy J, Kirkpatrick S, Hiza HAB, Kuczynski KJ, et al. Update of the healthy eating index: HEI-2010. J Acad Nutr Diet 2013; 113:569-80.^,1010. Guenther PM, Kirkpatrick S, Reedy J, Krebs-Smith SM, Buckman DW, Casavale KO, et al. The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans. J Nutr 2014; 144:399-407.. Considering that food intake is positively correlated with energy intake, the calculation of the number of portions was proportional to 1,000Kcal, similarly to the procedure adopted in the DQI-DFG-2013 ¹⁴14. Caivano S, Domene SMA. Diet quality index for healthy food choices. Rev Nutr 2013; 26:693-9.. The data in Table 3 show that Refined cereals is the group that resulted in the higher correlation with energy, however very weak (r = 0.24), suggesting that all the Index components are independent from the energy value of the menu. This result is similar to that observed for the application of the HEI-2010 (r = -0.21 for the component Fruit) ⁹9. Guenther PM, Casavale KO, Reedy J, Kirkpatrick S, Hiza HAB, Kuczynski KJ, et al. Update of the healthy eating index: HEI-2010. J Acad Nutr Diet 2013; 113:569-80.^,1010. Guenther PM, Kirkpatrick S, Reedy J, Krebs-Smith SM, Buckman DW, Casavale KO, et al. The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans. J Nutr 2014; 144:399-407.. The seven components that showed significant correlation with the energy resulted in negative coefficients, except for Refined cereals - the same result presented in the HEI-2010 ⁹9. Guenther PM, Casavale KO, Reedy J, Kirkpatrick S, Hiza HAB, Kuczynski KJ, et al. Update of the healthy eating index: HEI-2010. J Acad Nutr Diet 2013; 113:569-80.^,1010. Guenther PM, Kirkpatrick S, Reedy J, Krebs-Smith SM, Buckman DW, Casavale KO, et al. The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans. J Nutr 2014; 144:399-407..

Analysis of agreement between the experts’ opinion about the quality of menus and the result of the DQI-DFG-2019

a) Profile of the experts

Of the 67 experts selected and invited into the study, 26 (39%) agreed to participate, with 73% of them being affiliated with public institutions, regionally distributed as follows: 7% from the Northeast, 27% from the Central, 47% from the Southeast, and 19% from the South region of the country. All the researchers have worked on the theme of Healthy Eating for at least six years (46% for 6-10 years, 35% for 11-20 years, and 19% for 20-30 years); 35% of them classified their own experience on the subject as “proper qualification” and 65% as “highly qualified”. Regarding the theme Diet Quality Index, 27% report having experience above 10 years, 50% between five and nine years, 23% between two and four years; 27% classify their condition as “highly qualified”, 65% understand that they have “proper qualification”, and 7% declare themselves with “inappropriate qualification”. These data show homogeneity between the research participants regarding knowledge on the subjects under study.

b) Analysis of the menus through the classifications “good”, “regular” and “bad”

For the ten menus evaluated by the 26 experts, 84.6% of the classifications attributed agreed with the result presented by the DQI-DFG-2019. All the analyses resulted in high agreement, except for menu 5 (54% agreement).

c) Analysis of the menus through grades (0 to 10 points)

Table 4 shows the coefficients of the association between the score assigned by each expert in the ten menus and the result of the DQI-DFG-2019. The results obtained allow to claim that there is strong correlation in all situations studied.

Reliability

Reliability based on the evaluators

To determine if the Index produces the same score in each application and in identical situations there are some statistical tools available. As the result of the DQI-DFG, by definition, is similar for similar diets, when collected and registered in the same way, sources of error are attributed to the respondent, to the data collection, or to their processing. For this reason, we opted for reliability from the evaluators through preset menus for all of them. The agreement analysis between the result of the DQI-DFG-2019 and the result of each evaluator allows to conclude that they all agree with the result of the DQI-DFG-2019 (Table 5), and that the evaluators agree between themselves (correlation coefficient of 0.99 [CI: 0.98-1.00]; analysis of variance: p = 0.678).

Correlation between the components of the DQI-DFG and the total score

Since the diet may not satisfy the reference standard for all food groups, we conducted analysis of the correlation between each of the Index components and the total score to know which component has greater influence on the total score of the Index. Moderate and high correlations indicate greater influence on the final result. Therefore, the Index will present better results when it has proportional and similar correlations. This study presents correlation ranged from r = 0.10 for Refined cereals and r = 0.49 for Vegetables (Table 3), that is, weak or very weak correlations. Although the groups of Vegetables and Fruits are the groups with the highest correlation in comparison to the others (r = 0.49 and r = 0.48, respectively), both values are classified as weak correlation. This is a better result than that found in HEI-2010 ⁹9. Guenther PM, Casavale KO, Reedy J, Kirkpatrick S, Hiza HAB, Kuczynski KJ, et al. Update of the healthy eating index: HEI-2010. J Acad Nutr Diet 2013; 113:569-80.^,1010. Guenther PM, Kirkpatrick S, Reedy J, Krebs-Smith SM, Buckman DW, Casavale KO, et al. The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans. J Nutr 2014; 144:399-407., where seven components presented moderate correlations.

According to Cronbach ⁴³43. Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika 1951; 16:297-334., internal consistency analysis evaluates the degree of association between the internal components of the instrument through the means of the correlations between all possible combinations. Several authors, such as those responsible for the publication of the HEI ⁸8. Kennedy ET, Ohls J, Carlson S, Fleming K. The healthy eating index: design and applications. J Am Diet Assoc 1995; 95:1103-8.^,99. Guenther PM, Casavale KO, Reedy J, Kirkpatrick S, Hiza HAB, Kuczynski KJ, et al. Update of the healthy eating index: HEI-2010. J Acad Nutr Diet 2013; 113:569-80.^,1010. Guenther PM, Kirkpatrick S, Reedy J, Krebs-Smith SM, Buckman DW, Casavale KO, et al. The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans. J Nutr 2014; 144:399-407.^,1212. Guenther PM, Reedy J, Krebs-Smith SM. Development and evaluation of the Healthy Eating Index-2005: technical report. Washington DC: Center for Nutrition Policy and Promotion; 2007., use Cronbach’s alpha coefficient as a way to assess reliability through the internal consistency of the instrument while acknowledging its limitations ¹⁰10. Guenther PM, Kirkpatrick S, Reedy J, Krebs-Smith SM, Buckman DW, Casavale KO, et al. The Healthy Eating Index-2010 is a valid and reliable measure of diet quality according to the 2010 Dietary Guidelines for Americans. J Nutr 2014; 144:399-407.. However, this analysis presents weakness if applied to the DQI-DFG, since the internal components of this Index do not correlate. Applying this coefficient would be an assessment referring to the response pattern of the diets studied and not to a characteristic of the instrument, because the value of alpha will suffer modifications depending on the pattern of diets to which the Index is applied ⁴⁴44. Streiner DL. Being inconsistent about consistency: when coefficient alpha does and doesn't matter. J Pers Assess 2003; 80:217-22..

Limitations of the DQI-DFG

The limitations of this study are focused on the construction of the Index, that is, for its content validation. There are some concepts of dietary technique that need to be analyzed and contemplated in the Index in order to preserve the nutritional value of foods. It is possible to highlight some points: presence of fruit juices (There is loss of fibers and nutrients during preparation, and satiety is lower ⁶6. Departamento de Atenção Básica, Secretaria de Atenção à Saúde, Ministério da Saúde. Guia alimentar para a população brasileira. Brasília: Ministério da Saúde; 2014.), preparation technique, sources of organic food and division of meals throughout the day.

Final considerations

The DQI-DFG showed consistent validity and reliability characteristics according to the criteria adopted in this study. This Index is the first to submit the criteria established for its development to a consultation with experts in the field, which provided authentic results about the diet quality. Construct validation based on a reference diet is another element that differentiates the development of this index, either from national or international ones. Other dimensions of the DQI-DFG that also express the original contribution of this study include the identification of Legumes as an adequacy component, the non-inclusion of alcohol to obtain the total score, the composition of the components of analysis focused on food groups and not on nutrients, and the attribution of decreasing points when there is excess intake of Cereals, tubers and roots, Fruits, and Milk and dairy products. This assessment is based on the agreement between the experts’ opinion and the results generated by the Index, both in relation to the classification of the diet into categories and to the result of the total score. The reproducibility of the DQI-DFG was confirmed through its application by different evaluators.

Therefore, the results of this work contributed to the construction of the components of the DQI-DFG-2019 through parameters supported by scientific evidence, by the experts’ opinion, and by a diet analysis model that promotes the valuation of regional eating habits, considering that the simplicity of the moderation and adequacy components allow the classification of diverse eating patterns.

Maintaining a regular pattern with regard to healthy eating promotes the conditions for protection against chronic non-communicable diseases and deficiency diseases. Understanding diet quality through monitoring and evaluating food intake is a strategy that can promote healthy choices and assist in nutritional care. This Index assesses diet quality to facilitate the monitoring and evaluation of food intake, as well as to know individuals dietary patterns.

Thus, it is possible to apply this Index as a tool for food intake studies. Knowing the dietary patterns of the population promotes the development of health care and intervention actions as well as the definition of public policies. This Index can also be used as a food and nutrition education tool. When the individual can explore the characteristics of their diet through an indicator and composition in food groups, it is possible to generate autonomy and freedom of food choice. Thus, this tool can be suggested as a complement to the guidelines of the current Food Guide for the Brazilian Population.

Acknowledgments

We would like to thank you for giving away the data from ISA-Camp 2008.

References

Publication Dates

History