Validation of an instrument to assess food diversity in women of childbearing age in Medellín, Colombia - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Public Health Nutr . 2022 Apr 8;25(10):2709–2719. doi: 10.1017/S1368980022000854 Search in PMC Search in PubMed View in NLM Catalog Add to search Validation of an instrument to assess food diversity in women of childbearing age in Medellín, Colombia Nathalia Correa Guzmán Nathalia Correa Guzmán 1 School of Nutrition and Dietetics, University of Antioquia (Universidad de Antioquia), Medellín 051054, Colombia 2 Research Group on Food and Human Nutrition, Medellín, Colombia Find articles by Nathalia Correa Guzmán 1, 2, * , Víctor Daniel Calvo Betancur Víctor Daniel Calvo Betancur 1 School of Nutrition and Dietetics, University of Antioquia (Universidad de Antioquia), Medellín 051054, Colombia Find articles by Víctor Daniel Calvo Betancur 1 , Diana María Sepúlveda Herrera Diana María Sepúlveda Herrera 1 School of Nutrition and Dietetics, University of Antioquia (Universidad de Antioquia), Medellín 051054, Colombia Find articles by Diana María Sepúlveda Herrera 1 , Diana Liseth Cárdenas Sánchez Diana Liseth Cárdenas Sánchez 1 School of Nutrition and Dietetics, University of Antioquia (Universidad de Antioquia), Medellín 051054, Colombia 2 Research Group on Food and Human Nutrition, Medellín, Colombia Find articles by Diana Liseth Cárdenas Sánchez 1, 2 , Luz Mariela Manjarrés Correa Luz Mariela Manjarrés Correa 1 School of Nutrition and Dietetics, University of Antioquia (Universidad de Antioquia), Medellín 051054, Colombia 2 Research Group on Food and Human Nutrition, Medellín, Colombia Find articles by Luz Mariela Manjarrés Correa 1, 2 Author information Article notes Copyright and License information 1 School of Nutrition and Dietetics, University of Antioquia (Universidad de Antioquia), Medellín 051054, Colombia 2 Research Group on Food and Human Nutrition, Medellín, Colombia * Corresponding author : Email [email protected] Received 2021 Sep 17; Revised 2022 Mar 11; Accepted 2022 Apr 4; Collection date 2022 Oct. © The Authors 2022 This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike licence ( https://creativecommons.org/licenses/by-nc-sa/4.0/ ), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the same Creative Commons licence is included and the original work is properly cited. The written permission of Cambridge University Press must be obtained for commercial re-use. PMC Copyright notice PMCID: PMC9991835 PMID: 35393003 Abstract Objective: To validate a Food Diversity Questionnaire (CDA, for its name in Spanish) that identifies the prevalence of the risk of deficiency in the intake of eleven micronutrients. Design: The CDA paper form, an online application for data entry and handling, was designed and compared with the 24-h recall (24HR) as a reference method. All data were processed in Personal Computer Software for Intake Distribution Estimation (PC-SIDE) v1 software. A descriptive analysis and comparisons between prevalence, concordance and reproducibility analyses were performed. Setting: Medellín, Colombia. Participants: Women of childbearing age between 19 and 50 years ( n 186) who worked for the Buen Comienzo programme in 2019. Results: When comparing the adjusted 24HR technique and the CDA, there was no significant difference in population-level data at risk of deficiency in any micronutrient intake. However, based on individual-level data of the best linear unbiased predictor, the concordance analyses were weak, and although agreements were high according to the diagnostic performance tests, a good ability to detect deficiency was only observed in a few nutrients: vitamin A 100·0 %, Ca 98·7 %, Fe 92·8 %, folates 91·6 %, and pyridoxine 81·8 %. Conclusions: The CDA validated in this study is useful and faster at evaluating population-level data at risk of deficiency in the intake of Ca, Fe, Zn, thiamine, riboflavin, niacin, pyridoxine, folates, vitamin B 12 , vitamin C and vitamin A. Based on individual-level data, a good ability to detect deficiencies was observed in the intake of vitamin A, Ca, Fe, folates and pyridoxine. Keywords: Validation, Questionnaire, Food Diversity, 24-h recall, Women Throughout history, researchers have developed and perfected different methods for collecting information on dietary intake, which has been associated with eating habits, energy and nutrient consumption, and health and disease states. Some of the methods of application at the individual level are dietary history, dietary record, 24-h dietary recall (24HR) and frequency of food consumption ( 1 , 2 ) . In Colombia, some of these evaluation methods for dietary intake have been used in national surveys ( 2 – 4 ) , departmental surveys ( 5 ) and municipal surveys ( 6 ) . Most of these studies have applied 24HR, considered the most appropriate method to estimate usual dietary intake distributions and to calculate the prevalence of the risk of any energy or nutrient intake deficiency ( 7 ) . However, 24HR is an expensive and time-consuming method to administer and analyse. Currently, the academic environment and those responsible for developing public health policies are demanding new methods of evaluating dietary intake that are faster and less expensive ( 8 ) . Along these lines, the FAO of the UN ( 9 ) recommends applying the dietary (food) diversity method. The food diversity method recommended by the FAO is a proxy for the risk of nutritional deficiency ( 9 ) . They propose a qualitative technique similar to that applied in 24HR, in which they ask about food and beverages consumed during the last 24 h, but based on food groups, without determining the amount consumed. The analyses can be performed using scores calculated by adding the different consumed food groups or by eating patterns. Focusing on the food groups of interest, it can be applied at the household or individual level. The FAO recommends that the diversity form be adapted and validated in each population before applying ( 9 ) . Although different studies of food diversity have been conducted internationally for a couple of decades ( 10 – 13 ) , they usually assess diet in a general way and categorise individuals according to whether their eating behaviour is considered healthy; they do not predict disease or mortality but rather measure adherence to dietary guidelines. In addition, they have used different collection instruments and have developed various methods that vary according to the objectives of the researcher ( 14 ) because there is still no consensus on the form to be used or on how to define the minimum amount of food eaten. They generally use 15 g as a cut-off to count as the consumption of a food group ( 9 , 15 – 17 ) . Taking into account two limitations – the first that a validated diversity form is not available for women of childbearing age in the Colombian population and the second that without estimating the amount of food consumed, the nutrient contribution cannot be calculated, and the prevalence of the risk of a deficiency thus cannot be calculated – we asked if a Food Diversity Questionnaire (CDA, for its name in Spanish) that estimated the amount of food consumed could serve as a proxy for the risk of deficiency in the intake of eleven micronutrients, similar to that obtained using the 24HR method. Thus, this study aimed to validate a CDA that identified the prevalence of the risk of deficiency in the intake of eleven micronutrients in women of childbearing age. Methods Type of study This is an observational, descriptive, cross-sectional epidemiological study and validation of eleven micronutrient intakes by the CDA compared with 24HR. Population Women of childbearing age between 19 and 50 years of who worked for the Buen Comienzo (Good Start) programme of the city of Medellín, Colombia, in 2019. Buen Comienzo is a programme initiated by the municipal government of Medellín, Colombia, which, through different types of care, provides early education to families and children for their first 5 years. The educational agents of the programme are mainly women ( 18 ) . Sample A total of 186 women were selected by probabilistic sampling through valid scientific inference and not by population representativeness ( 19 ) . The study is an analysis of two independent methods and moments – the 24HR and the CDA – applied to the same women with an interval of approximately 1 to 2 months between each method. Stata 15 software was used to run Fisher’s z-test to compare two independent correlations following the methods of Arimond et al. ( 15 ) , who analysed women of reproductive age in Bangladesh, disaggregated into twenty-one food groups and with a minimum inclusion of 15 g: the correlation of 24HR was 0·42, and a correlation of 0·7 was assumed for the CDA. The parameters for the sample calculation included a type I error of 0·05, a type II error of 0·20 and an allocation ratio (n 2 /n 1 ) = 1. A two-tailed hypothesis was set with a CI of 95 %. The Technical Directorate of the Buen Comienzo programme authorised the study at fifteen of its centres which were randomly selected, and the questionnaires were administered to all the assistants and teachers in the centres until we met the estimated sample number. Nine centres participated in the study in total. In these 9 centres, 191 participants were approached; those who were on vacation, sick leave or in other activities were contacted three more times to check on their participation; those who decided not to participate or were excluded were replaced by the next person on the list at the centre. Based on the selection criteria, two men, a lactating woman and a pregnant woman were excluded, and one additional woman decided not to participate. Selection criteria Inclusion criteria Women between 19 and 50 years of age agreed to participate in this study and worked in the selected centres of the Buen Comienzo programme in Medellín, Colombia, in 2019. Exclusion criteria Women in the period of gestation or lactation or with a diagnosis of pathologies that affect feeding, such as diabetes, celiac disease and dyslipidaemia. Data collection The survey schedule was carried out according to each participant. The surveys were applied in 2-d intervals to ensure that they were not administered on consecutive days and were distributed on different days of the week. The surveys were applied during working weekdays and at the homes of the participants on weekends. Food was not provided by the institution. The four interviewers and four data entry clerks were dietitian nutritionists trained in the following techniques: Anthropometric measurements The interviewers were trained in the appropriate techniques for taking anthropometric measurements of weight and height, which were taken in the first interview with a digital scale with a capacity of 120 kg and precision of 100 g and a body height rod with a capacity of 2 m and sensitivity of 1 mm. Data were necessary to classify nutritional status according to BMI kg/m 2 in accordance with the values proposed by the WHO, to identify underweight women (<18·5), those with a normal BMI (≥ 18·5 to <25), those who are overweight (≥ 25 to <30) and those who are obese (≥ 30) ( 20 ) . 24-h food recall 24HR was the reference method to calculate the prevalence of the risk of deficiency in the usual intake of nutrients. The adjusted multistep technique was applied ( 21 ) , and the information was recorded on a paper form that detailed the preparations, the names of the foods, beverages, supplements, and complements, and the amount consumed by the respondent during the 24 h before the survey ( 7 ) . The 24HR survey took approximately 20 min to administer. In this study, each woman was given a minimum of five and a maximum of seven 24HR distributed throughout the days of the week on non-consecutive days, a procedure that was necessary to adjust intra- and interindividual variability ( 22 ) . To measure the amount consumed, a set of food models, geometric figures and a photo album with life-size utensils were used, all coded and tested in Colombia ( 23 , 24 ) . Some dichotomous verification questions and a space for noting useful observations were included. 24HR was entered into the Dietary Intake Evaluation software (Evindi v5) of the School of Nutrition and Dietetics of the University of Antioquia ( 25 ) . This software calculates the nutrients consumed in each of the 24HR from different food composition table ( 26 – 32 ) labels, supplements and preparations compiled in a database. The software does not allow blank spaces because doing so would overestimate the risk of deficiency in the intake of energy and nutrients. Food Diversity Questionnaire The CDA was the test method. As mentioned above, there is no validated CDA for women of childbearing age in the Colombian population, nor could we find forms that defined the amount of food eaten. For these reasons, we designed a survey involving the following steps: Selection of the estimated average nutrient requirement The estimated average requirement (EAR) of the energy and nutrient intake recommendations (RIEN) for the Colombian population ( 33 ) was taken as the reference value for the micronutrients of greatest interest in women of childbearing age: Ca (EAR 800 mg), Fe (EAR 11·7 mg), Zn (EAR 6·50 mg), vitamin A (EAR 500 retinol equivalents (RE)), thiamine (EAR 0·9 mg), riboflavin (EAR 0·9 mg), niacin (EAR 11 mg), pyridoxine (EAR 1·1 mg), folates (EAR 320 µg of dietary folate equivalents), vitamin B 12 (EAR 2·0 µg) and vitamin C (EAR 60 mg). Definition of food groups First, the source food groups of the selected micronutrients were identified, either by their high concentration of each nutrient or by a frequency and amount of consumption that made them a nutrient source in the Colombian population (Supplementary Material 1 ). Subsequently, foods for which 100 g ( 34 ) had a value greater than or equal to 10 % of the EAR of the selected micronutrients were identified so that these did not lead us to overestimate the micronutrient intake of each group; they were foods usually consumed according to the Food and Nutritional Security Profile of Medellín ( 6 ) . All foods within each group that had similar nutrients were grouped together. For example, the group including fruit was subdivided into two groups: the first with fruit rich in vitamin A and the second with fruit rich in vitamin C. In turn, each of these two groups was subdivided into subgroups that had a similar form of consumption, as explained below. Definition of food subgroups To quantify the amount consumed by the food group, all foods within each group that had a similar form of consumption were grouped, defining several subgroups. For example, the group including fruit as a source of vitamin C was subdivided into three subgroups: the first with fruits in the form of small sphere shapes, the second with fruits in the form of medium sphere shapes and the third showing figures representing the volume of fruits consumed in pieces or that have an irregular shape. In turn, each of these three subgroups was subdivided to measure them by glasses, mugs and cups when consumed as juice. Importantly, 100 ml of juice from any subgroup represents 25 % of the micronutrients of the fruits of the subgroup ( 25 ) . To facilitate the collection of data by the interviewers and to avoid having to resort to memory, a codebook was designed that showed the food models established by subgroup. Each life-size model, figure or photograph established for each subgroup has several codes representing different quantities (Supplementary Material 2 ). Standardisation of weights and measures To measure the amount of each food subgroup consumed, food models, geometric figures and photographs with life-size utensils coded and tested in Colombia were used ( 23 , 24 ) . Each of the foods was prepared and compared with the form that best represented it, and this amount was weighed three times to establish an average of each food per model. Finally, the average of the foods of each subgroup was calculated. Format of the Food Diversity Questionnaire A pilot study was performed to develop and design the format of the CDA. A total of thirty-five questionnaires were administered to women of childbearing age between 19 and 50 years who were conveniently selected to participate in this pilot and did not participate in the main study. Five different versions of the CDA were designed and tested to establish the version that best facilitated the recall of the respondents and completion by the interviewer. The CDA was selected to prevent the interviewer repeating questions, writing the same thing several times and looking at several pages to ask and write the answers. In the final format, the first side of the questionnaire covered identification data, control data for statistical adjustments, including the questionnaire number and day of week, useful notes for entering, and verification questions that also included the consumption and quantification of supplements and complements. The other side of the form covered subgroups and/or foods, groups, types of food, codes and quantities. To fill out this last part of the questionnaire, the first mealtime consumed the previous day was noted in the first row ‘Type of food’, and going down the form, all the food and/or drinks consumed at the mealtime were written, placing them in the corresponding subgroup, until all the foods consumed the previous day were listed. Lastly, the code representing each subgroup and the amount consumed the previous day in integer and/or decimal form (Supplementary Material 1 ) were recorded. If a number of different kinds of foods in the same subgroup were consumed, the interviewee was required to condense the foods into a single amount corresponding to the subgroup. Application of the questionnaire After the women answered the 24HR, it took between 1 and 2 months for the same women to receive at least one and at most two CDA distributed during the week on non-consecutive days to adjust the intra- and interindividual variability ( 22 ) . The CDA took approximately 10 min to fill out. Data processing To enter information for the CDA, an online application was designed that contained the same database of nutritional information as Evindi v5 ( 25 ) . From an administrative perspective, the online application allowed us to select the foods that made up each subgroup and to modify or enter nutritional information on foods, supplements and complements. For entering information in each of the surveys, all the items of the questionnaire appeared as tabs in the online application: identification, control data, CDA and questions. In the CDA with the list of supplements, complements, groups and subgroups of food, only the codes and quantities consumed were selected, without disaggregating by type of food as in the paper format. To generate the report, the application averaged the micronutrients of the foods of each subgroup. This average was multiplied by the code and the amount consumed by subgroup in each questionnaire. Then, the micronutrients of all subgroups, supplements and complements consumed according to the questionnaire were added. Finally, the micronutrient report for each individual was obtained from the questionnaire. Statistical analysis The nutrient database generated in Evindi v5 for the 24HR and the database with the micronutrients of each individual recorded by the CDA were migrated and processed in Personal Computer Software for Intake Distribution Estimation (PC-SIDE) v1 of Iowa State University ( 35 ) . This software estimates the distribution of the usual nutrient intake, calculates the proportion of the population at risk of deficiency in the consumption of nutrients from the EAR according to the RIEN for the Colombian population ( 33 ) and calculates the best linear unbiased predictor (BLUP), which is an approximation of the usual intake of each nutrient per individual ( 35 ) . All analyses in PC-SIDE were adjusted with a type I error of 0·15 according to Anderson and Darling ( 36 ) . In the descriptive analysis, summary indicators such as the arithmetic mean and standard deviation were used. To compare the adjusted prevalence of the risk of deficiency in the usual intake of micronutrients between the 24HR and CDA techniques (24HR refers to the adjustment of the five or seven 24HR and CDA refers to the adjustment of the two CDA), the crude standard error (SEc), the adjusted standard error (SEa) of the PC-SIDE v1 software, the 95 % CI calculated with the SEa and the proportional difference test with the adjusted prevalence of deficiency were calculated. The McNemar test was applied to compare the unadjusted prevalences between the first and second CDA (CDA1 refers to the crude first CDA and CDA2 refers to the crude second CDA), and the Pearson chi-squared test of independence was used to compare the unadjusted prevalences between the 24HR and CDA techniques (24HR refers to the crude first 24HR and CDA refers to the crude first CDA). For the concordance analyses between methods (the methods refer to the adjustment of the two CDA and to the adjustment of the five or seven 24HR) and for the reproducibility analyses between measurements (the measurements refer to the crude first CDA and to the crude second CDA), the intraclass correlation coefficient (ICC) was calculated for continuous variables, and Cohen’s kappa index was calculated for categorical variables. The diagnostic performance was compared between the adjustment of the two CDA and the adjustment of the five or seven 24HR and between the crude first CDA and the crude second CDA. The diagnostic performance was evaluated by its sensitivity, specificity, predictive value, likelihood ratio, entropy reduction and bias index. For all two-sided tests, a P -value of less than 0·05 was considered statistically significant. The data processing and analysis were performed in SPSS, Stata and OpenEpi software. Controlling for biases To control for selection biases, we ensured that the participation of the women was not influenced by the researchers or interviewers and was carried out according to the checklist established with the selection criteria, sampling processes and data collection. To control observer biases, the interviewers and data entry clerks were trained and supervised, and we reviewed the quality of the data. To control information biases, life-size figures, models and photographs were used to quantify food intake. For the control of random biases, since the intake varies unpredictably, between five and seven 24HR and between one and two CDA were given to each woman to adjust the intra- and interindividual variability by the number of questionnaires and days in the week in the PC-SIDE software. To control observer bias and prevent dropout, we implemented strategies to facilitate visits, agreeing on a schedule with each woman, and visited them at work during the week and at their homes during the weekend. Results Characterisation For each of the 186 women, the surveys were distributed on different days of the week. A total of 1122 24HR were submitted, for an average of six 24HR per person (at least five and at most seven 24HR). A total of 337 CDA were submitted (186 with the first questionnaire and 151 with the second questionnaire). The women had an average age of 32 years (7 sd ) and a BMI of 25·5 kg/m 2 (4·0 sd ), distributed as 1 % underweight, 49 % normal BMI, 37 % overweight and 13 % obese. Comparison between the 24-h recall and Food Diversity Questionnaire results Prevalence of adjusted risk of deficiency The prevalence of the risk of deficiency in women by the 24HR was approximately 70 % for the micronutrients of Ca, Fe and folates. In the diversity questionnaire, the prevalence of the risk of deficiency was higher (Table 1 ). When comparing the adjusted prevalences between the CDA and 24HR, no significant differences were found for any of the nutrients, for example, vitamin C ( P = 0·6071), folate ( P = 0·4667), Zn ( P = 0·4524), niacin ( P = 0·3703), Ca ( P = 0·3533) and Fe ( P = 0·3391) (Table 1 ). Table 1. Adjusted prevalence of the risk of deficiency in the usual intake of micronutrients by 24HR and the present CDA ( n 186) Nutrient 24HR * CDA * Crude P -value ‡ Adjusted P -value Adjusted prevalence of deficiency Adjusted prevalence of deficiency % 95 % CI † SEc ‡ SEa § % 95 % CI † SEc ‡ SEa § Ca 78·5 69·4, 87·6 0·0301 0·0465 91·0 78·5, 103·5 0·0210 0·0637 0·0008 0·3533 Fe 61·4 51·2, 71·6 0·0357 0·0520 75·9 62·2, 89·6 0·0314 0·0701 0·0026 0·3391 Zn 17·3 7·5, 27·0 0·0277 0·0497 21·6 6·0, 37·2 0·0302 0·0798 0·2948 0·4524 Vitamin A 4·2 0·0, 10·0 0·0147 0·0297 46·2 38·0, 54·3 0·0366 0·0417 <0·0001 0·0580 Thiamine 26·9 16·0, 37·9 0·0325 0·0558 45·2 35·6, 54·8 0·0365 0·0491 0·0002 0·2860 Riboflavin 4·2 0·0, 8·8 0·0147 0·0236 16·6 4·5, 28·7 0·0273 0·0616 0·0001 0·3355 Niacin 20·2 9·7, 30·8 0·0294 0·0538 32·4 0·0, 48·4 0·0343 0·0819 0·0075 0·3703 Pyridoxine 10·5 1·5, 19·4 0·0225 0·0456 33·7 23·2, 44·1 0·0347 0·0533 <0·0001 0·2303 Folate 74·9 64·8, 85·0 0·0318 0·0515 77·7 65·8, 89·6 0·0305 0·0608 0·5254 0·4667 Vitamin B 12 3·9 0·0, 9·5 0·0142 0·0287 16·5 2·7, 30·2 0·0272 0·0700 0·0001 0·3442 Vitamin C 25·5 15·0, 36·0 0·0320 0·0536 16·4 4·9, 27·9 0·0271 0·0586 0·0310 0·6071 Open in a new tab 24HR, 24-h recall; CDA, Food Diversity Questionnaire; SEc, crude standard error; Sea, adjusted standard error. * 24HR refers to the adjustment of the five or seven 24HR and CDA refers to the adjustment of the two CDA, adjusted in the Personal Computer Software for Intake Distribution Estimation (PC-SIDE) v1 ( 30 ) by number of questionnaires with a type I error of 0·15 according to Anderson and Darling ( 36 ) . † Calculated with the SEa. ‡ SEc and crude P -value was added to look at differences, but SEa and adjusted P -value were analysed. § Calculated in PC-SIDE v1. Concordance between the 24-h recall and the Food Diversity Questionnaire The BLUP was obtained for the concordance analyses between the 24HR questionnaire and the CDA for each individual. The ICC and Cohen’s kappa that measure the agreement between the 24HR and CDA on all micronutrients were weak. However, there were high percentages of agreement that possibly reflected the ability of the CDA to distinguish a subject with micronutrient deficiency from a subject without micronutrient deficiency (Table 2 ). Table 2. Concordance between the 24HR and the CDA ( n 186) Nutrient Variability between methods * Continuous measurement Categorical measurement ICC † 95 % CI P -value Kappa 95 % CI P -value % Agreement ‡ Ca 0·383 § 0·254, 0·499 <0·0001 0·100 −0·029, 0·230 0·0114 80·7 Fe 0·112 § −0·032, 0·252 0·0630 0·187 0·054, 0·321 0·0011 69·9 Zn 0·394 § 0·265, 0·508 <0·0001 0·275 0·060, 0·490 0·0001 87·6 Vitamin A 0·361 § 0·229, 0·479 <0·0001 0·026 −0·010, 0·062 0·0586 55·9 Thiamine 0·066 § −0·079, 0·207 0·1860 0·050 −0·079, 0·178 0·2221 55·9 Riboflavin 0·401 § 0·273, 0·515 <0·0001 −0·019 −0·046, 0·007 0·6632 90·9 Niacin 0·362 § 0·230, 0·480 <0·0001 0·069 −0·084, 0·223 0·1665 73·7 Pyridoxine 0·248 § 0·109, 0·378 <0·0001 0·198 0·071, 0·325 <0·0001 74·7 Folate 0·418 § −0·292, 0·530 <0·0001 0·288 0·109, 0·467 <0·0001 81·7 Vitamin B 12 0·247 § 0·107, 0·377 <0·0001 0·110 −0·122, 0·342 0·0280 93·0 Vitamin C 0·485 § −0·367, 0·588 <0·0001 0·179 0·023, 0·336 0·0024 78·0 Open in a new tab 24HR, 24-h recall; CDA, Food Diversity Questionnaire; ICC, intraclass correlation coefficient. * The methods refer to the adjustment of the two CDA and to the adjustment of the five or seven 24HR. † Type C ICC that use a definition of coherence. The variance in the intermediate measure is excluded from the variance in the denominator. ‡ The agreement or comparison between two methods on the same sample ( 43 ) . § The estimator is the same whether the interaction effect is present or not. Diagnostic performance tests of the Food Diversity Questionnaire For performance tests between the 24HR questionnaire and the CDA, the BLUP was obtained for each individual. The CDA showed a high sensitivity to detect individuals deficient in vitamin A (100·0 %), Ca (98·7 %), Fe (92·8 %), folates (91·6 %) and pyridoxine (81·8 %). According to the reduction in entropy after a positive test, two micronutrients with high efficacy were observed: folates (7·3 %) and Fe (3·2 %). In other words, for example, the CDA was 1·4 times more likely to return a positive result in individuals with folate deficiency than in those without folate deficiency (Table 3 ). Table 3. Diagnostic performance of the CDA compared to 24HR ( n 186) Nutrient * Sensitivity 95 % CI Specificity 95 % CI Predictive value Likelihood ratio Reduction of entropy Bias index Positive 95 % CI Negative 95 % CI Positive 95 % CI Negative 95 % CI Positive% Negative% Ca 98·7 95·2, 99·6 † 8·1 2·8, 21·3 † 81·2 74·9, 86·2 † 60·0 23·1, 88·2 † 1·074 1·013, 1·138 0·166 0·000, 724·800 1·6 −17·4 0·1720 Fe 92·8 86·9, 96·2 † 23·0 14·2, 34·9 † 71·2 63·8, 77·6 † 60·9 40·8, 77·8 † 1·204 1·154, 1·257 0·314 0·158, 0·624 3·2 −3·7 0·2043 Zn 31·6 15·4, 54·0 † 94·0 89·3, 96·7 † 37·5 18·5, 61·4 † 92·4 87·4, 95·5 † 5·274 2·136, 13·020 0·728 0·625, 0·847 −33·2 6·0 −0·0161 Vitamin A 100·0 34·2, 100·0 † 55·4 48·2, 62·4 † 2·4 0·7, 8·3 † 100·0 96·4, 100·0 † 2·244 2·191, 2·298 0·000 0·000, 0·000 −5·3 0·0 0·4409 Thiamine 48·8 34·3, 63·5 † 57·9 49·8, 65·7 † 24·7 16·6, 35·1 † 80·0 71·4, 86·5 † 1·160 1·013, 1·327 0·884 0·792, 0·987 −3·2 2·7 0·2151 Riboflavin 0·0 0·0, 65·8 † 91·9 87·0, 95·0 † 0·0 0·0, 20·4 † 98·8 95·8, 99·7 † 0·000 0·000, 0·000 1·089 0·000, 0·000 0·0 −0·4 0·0699 Niacin 26·9 13·7, 46·1 † 81·3 74·5, 86·5 † 18·9 9·5, 34·2 † 87·3 80·9, 91·7 † 1·436 0·629, 3·278 0·899 0·808, 1·001 −8·1 2·3 0·0591 Pyridoxine 81·8 52·3, 94·9 † 74·3 67·3, 80·2 † 16·7 9·0, 28·7 † 98·5 94·6, 99·6 † 3·182 2·902, 3·488 0·245 0·091, 0·656 −22·6 14·6 0·2312 Folate 91·6 86·1, 95·0 † 34·4 20·4, 51·7 † 87·0 81·0, 91·4 † 45·8 27·9, 64·9 † 1·395 1·269, 1·534 0·246 0·150, 0·401 7·3 −23·1 0·0430 Vitamin B 12 33·3 6·2, 79·2 † 94·0 89·6, 96·6 † 8·3 1·5, 35·4 † 98·85 95·91, 99·68 † 5·545 0·092, 334·000 0·709 0·266, 1·891 −20·4 2·0 0·0484 Vitamin C 19·5 10·2, 34·0 † 94·5 89·5, 97·2 † 50·0 28·0, 72·0 † 80·6 74·0, 85·8 † 3·537 1·008, 12·410 0·852 0·802, 0·905 −16·6 3·5 −0·1344 Open in a new tab 24HR, 24-h recall; CDA, Food Diversity Questionnaire. * The diagnostic performance was compared between the adjustment of the two CDA and the adjustment of the five or seven 24HR. † Method: Wilson points. Intratechnique analysis of the Food Diversity Questionnaire Unadjusted prevalence of risk of deficiency When comparing the unadjusted prevalences between the first and second CDA, no significant differences were found for any nutrients, except for vitamin C, although the CI of vitamin C at some point intersected. On the other hand, when comparing the unadjusted prevalences between CDA and 24HR, statistically significant differences were found for all nutrients except Zn and vitamin B 12 (Table 4 ). Table 4. Unadjusted prevalence of risk of deficiency in the usual nutrient intake Nutrient Unadjusted prevalence of deficiency ( n 151) Unadjusted prevalence of deficiency ( n 186) CDA1 * CDA2 * P -value † 24HR * CDA * P -value ‡ % 95 % CI % 95 % CI % 95 % CI % 95 % CI Ca 81·5 75·2, 87·7 83·4 77·5, 89·4 0·7200 74·2 71·7, 76·8 82·2 78·1, 86·3 0·0027 Fe 79·5 73·0, 86·0 83·4 77·5, 89·4 0·3450 67·7 65·0, 70·5 81·3 77·1, 85·5 <0·0001 Zn 40·4 32·5, 48·3 30·5 23·0, 37·9 0·0860 38·1 35·3, 40·1 35·0 29·9, 40·1 0·2992 Vitamin A 53·0 44·9, 61·0 53·0 44·9, 61·0 1·0000 35·7 32·8, 38·5 53·7 48·4, 59·1 <0·0001 Thiamine 54·3 46·3, 62·3 55·6 47·6, 63·6 0·8990 42·0 39·1, 44·9 53·1 47·8, 58·5 0·0003 Riboflavin 33·1 25·5, 40·7 26·5 19·4, 33·6 0·1930 19·1 16·8, 21·4 30·0 25·1, 34·9 <0·0001 Niacin 48·3 40·3, 56·4 52·3 44·3, 60·4 0·5390 35·4 32·6, 38·2 49·9 44·5, 55·2 <0·0001 Pyridoxine 44·4 36·4, 52·4 45·7 37·7, 53·7 0·8940 32·8 30·1, 35·6 44·5 39·2, 49·8 <0·0001 Folate 79·5 73·0, 86·0 84·8 79·0, 90·6 0·2150 75·2 72·7, 77·8 81·9 77·8, 86·0 0·0109 Vitamin B 12 35·8 28·0, 43·5 29·8 22·4, 37·2 0·2720 28·0 25·4, 30·6 32·0 27·0, 37·1 0·1494 Vitamin C 27·2 20·0, 34·3 38·4 30·6, 46·3 0·0220 47·7 44·8, 50·6 33·2 28·2, 38·3 <0·0001 Open in a new tab 24HR, 24-h recall; CDA, Food Diversity Questionnaire; CDA1, first Food Diversity Questionnaire; CDA2, second Food Diversity Questionnaire. * CDA1 refers to the crude first CDA, CDA2 refers to the crude second CDA, 24HR refers to the crude first 24HR and CDA refers to the crude first CDA. † Based on the McNemar test. ‡ Based on the chi-squared test of independence. Reproducibility between the first and second Food Diversity Questionnaires To analyse the reproducibility between the first and second unadjusted CDA, the risk of deficiency for each nutrient was classified in each questionnaire. The ICC and Cohen’s kappa measuring the agreement between the first and second CDA were weak for each micronutrient. However, there were high percentages of agreement that possibly reflected the ability of the first CDA to distinguish a subject with micronutrient deficiency from a subject without micronutrient deficiency (Table 5 ). Table 5. Reproducibility between the first and second CDA ( n 151) Nutrient Variability between measurements * Continuous measurement Categorical measurement ICC † 95 % CI P -value Kappa 95 % CI P -value % agreement ‡ Ca 0·313 § 0·161, 0·449 <0·0001 0·291 0·101, 0·481 0·0002 79·5 Fe 0·397 § 0·254, 0·523 <0·0001 0·388 0·204, 0·572 <0·0001 81·5 Zn 0·327 § 0·177, 0·463 <0·0001 0·041 −0·116, 0·197 0·3048 55·6 Vitamin A 0·243 § 0·087, 0·387 0·0010 0·282 0·129, 0·435 0·0003 64·2 Thiamine 0·487 § 0·355, 0·600 <0·0001 0·171 0·013, 0·328 0·0179 58·9 Riboflavin 0·469 § 0·335, 0·585 <0·0001 0·244 0·082, 0·407 0·0012 68·2 Niacin 0·352 § 0·205, 0·484 <0·0001 0·127 −0·030, 0·285 0·0585 56·3 Pyridoxine 0·391 § 0·247, 0·518 <0·0001 0·251 0·096, 0·406 0·0010 62·9 Folate 0·494 § 0·363, 0·605 <0·0001 0·282 0·095, 0·469 0·0002 78·8 Vitamin B 12 0·127 § −0·033, 0·280 0·0600 0·207 0·046, 0·368 0·0052 64·9 Vitamin C 0·336 § 0·187, 0·471 <0·0001 0·274 0·119, 0·429 0·0003 67·6 Open in a new tab CDA, Food Diversity Questionnaire; ICC, intraclass correlation coefficient. * The measurements refer to the crude first CDA and to the crude second CDA. † Type C intraclass correlation coefficients that use a definition of coherence. The variance in the intermediate measure is excluded from the variance in the denominator. ‡ The agreement or comparison between two measurements on the same samples ( 43 ) . § The estimator is the same whether the interaction effect is present or not. Diagnostic performance of the first Food Diversity Questionnaire To compare the performance of the unadjusted first and second CDA, the risk of deficiency of each nutrient was classified in each questionnaire. The second CDA showed a high sensitivity for detecting individuals deficient in Fe (90·8 %), folates (90·0 %), Ca (88·6 %), vitamin A (66·3 %) and thiamine (63·4 %). According to the reduction in entropy after a positive test, four micronutrients for which CDA had high efficacy were observed: Fe (11·2 %), Ca (8·4 %), folate (7·4 %) and vitamin A (5·2 %). In other words, for example, it was 1·7 times more likely that the CDA returned a positive result in individuals with an Fe deficiency than in those without Fe deficiency (Table 6 ). Table 6. Diagnostic performance of the second CDA with the first as a reference ( n 151) Nutrient * Sensitivity 95 % CI Specificity 95 % CI Predictive value Likelihood ratio Reduction of entropy Bias index Positive 95 % CI Negative 95 % CI Positive 95 % CI Negative 95 % CI Positive% Negative% Ca 88·6 81·8, 93·1 † 39·3 23·6, 57·6 † 86·5 79·5, 91·4 † 44·0 26·7, 62·9 † 1·460 1·298, 1·642 0·290 0·191, 0·439 8·4 −20·6 0·0199 Fe 90·8 84·3, 94·8 † 45·2 29·2, 62·2 † 86·5 79·5, 91·4 † 56·0 37·1, 73·3 † 1·656 1·473, 1·862 0·203 0·143, 0·288 11·2 −17·8 0·0397 Zn 32·8 22·3, 45·3 † 71·1 61·0, 79·5 † 43·5 30·2, 57·8 † 61·0 51·4, 69·7 † 1·135 0·861, 1·496 0·945 0·890, 1·004 −1·0 0·6 −0·0993 Vitamin A 66·3 55·4, 75·7 † 62·0 50·3, 72·4 † 66·3 55·4, 75·7 † 62·0 50·3, 72·4 † 1·742 1·590, 1·909 0·545 0·493, 0·602 5·2 2·7 0·0000 Thiamine 63·4 52·6, 73·0 † 53·6 42·0, 64·9 † 61·9 51·2, 71·6 † 55·2 43·4, 66·5 † 1·367 1·258, 1·486 0·682 0·611, 0·763 2·5 0·2 0·0133 Riboflavin 42·0 29·4, 55·8 † 81·2 72·5, 87·6 † 52·5 37·5, 67·1 † 73·9 65·0, 81·2 † 2·233 1·770, 2·816 0·714 0·664, 0·769 −5·7 6·1 −0·0662 Niacin 58·9 47·5, 69·5 † 54·9 42·9, 64·5 † 54·4 43·5, 65·0 † 58·3 46·8, 69·0 † 1·276 1·171, 1·391 0·763 0·687, 0·848 0·3 1·3 0·0397 Pyridoxine 59·7 47·7, 70·6 † 65·5 54·8, 74·8 † 58·0 46·2, 68·9 † 67·1 56·3, 76·3 † 1·729 1·564, 1·912 0·616 0·562, 0·674 0·6 5·3 0·0133 Folate 90·0 83·3, 94·2 † 35·5 21·1, 53·1 † 84·4 77·1, 89·7 † 47·8 29·2, 67·0 † 1·395 1·262, 1·542 0·282 0·173, 0·459 7·4 −18·5 0·0530 Vitamin B 12 42·6 30·3, 55·8 † 77·3 68·0, 84·5 † 51·1 37·0, 65·0 † 70·8 61·5, 78·6 † 1·878 1·531, 2·303 0·743 0·692, 0·797 −4·1 4·8 −0·0596 Vitamin C 61·0 45·7, 74·3 † 70·0 60·9, 77·8 † 43·1 31·2, 55·9 † 82·8 73·9, 89·1 † 2·033 1·822, 2·268 0·558 0·488, 0·637 −9·9 12·6 0·1126 Open in a new tab CDA, Food Diversity Questionnaire. * The diagnostic performance was compared between the crude first CDA and to the crude second CDA. † Method: Wilson points. Discussion In this study, according to the comparisons between methods with statistical adjustment in PC-SIDE v1, the CDA was useful for detecting the prevalence of micronutrient deficiency in the population, as we did not find statistically significant differences in any micronutrient between the CDA and 24HR. However, it was not useful for detecting the individual prevalence via the BLUP, since the concordance analyses were weak, and although the agreements were high according to the diagnostic performance tests, only a good ability to detect a deficiency in some micronutrients was observed: vitamin A (100·0 %), Ca (98·7 %), Fe (92·8 %), folates (91·6 %) and pyridoxine (81·8 %). The CDA without statistical adjustment was not useful for detecting the prevalence of micronutrient deficiency in the population, although in the intramethod analysis between the first and second CDA without statistical adjustment, there were no significant differences in the prevalence of almost all micronutrients. When comparing the prevalences between the 24HR and CDA methods without statistical adjustment, there were statistically significant differences in almost all micronutrients. Likewise, the reproducibility analyses were weak, and although the agreements were high according to the diagnostic performance tests, only a good ability to detect deficiency in some micronutrients was observed: Fe (90·8 %), folates (90·0 %), Ca (88·6 %), vitamin A (66·3 %) and thiamine (63·4 %). According to the above, it is necessary to use food models that quantify the amount consumed for the survey to be valid, to apply two questionnaires of food diversity on non-consecutive days and to send the data to PC-SIDE v1 to perform the statistical adjustment. The CDA validated in this study, although differing in methodology from other studies ( 37 ) , yielded results similar to those from studies in Mali, Mozambique, Bangladesh, Burkina Faso and the Philippines. Those studies, aiming to evaluate diversity indicators as a proxy for the adequacy of micronutrients at the population level, used 24HR and found that eight established food groups were correlated with the mean probability of adequacy, and the correlations were higher with higher levels of food group disaggregation and with the 15-g minimum requirement ( 15 ) . The reviewed studies that evaluated and validated CDA did not use the methods described in this study, mainly because they based their analyses on qualitative measures without quantifying the amount of food consumed ( 38 ) . In addition, most of them compared dependent techniques, that is, they built the reference and test indicators from the same instrument ( 16 ) . In this study, with a time interval between the application of both techniques, two independent methods were applied to the same women: 1–2 CDA as the test method and 5–7 24HR (to obtain a better fit) as the reference method. Although each study analysed food diversity differently, they almost all agreed on the food groups. The thirteen food groups and twenty-six food subgroups of this study are similar to those validated in the indicator for infants and young children, which includes seven groups: grains, roots and tubers; legumes and nuts; dairy products; meats; eggs; fruits and vegetables rich in vitamin A; and other fruits and vegetables ( 39 ) . They are also similar to groups used in the indicator of women’s dietary diversity ( 17 ) that includes these same seven groups but disaggregates them into different levels to yield twenty-one subgroups. Regarding the foods belonging to the groupings in the studies reviewed, most studies, including this one, incorporated only natural foods ( 40 ) . It is not clear whether ultraprocessed foods should be included or excluded, as some studies exclude, for example, embutidos (cured and dry sausages), fast food, packaged soups, packaged products and sweetened drinks ( 16 ) . In addition, some studies do not consider fortified foods, and most exclude supplements and complementary foods ( 41 ) , unlike this study, which included and quantified supplements and complementary foods since they provide significant amounts of nutrients. This study was not designed to evaluate the intake of calories, carbohydrates or fats; therefore, foods with high content of these nutrients were excluded, and the present CDA should not be used to measure their intake. In addition, although no validation tests were performed on the intake of protein or fibre, it would be worth performing these analyses because the food groups of the questionnaire include food sources of protein and fibre, and the questionnaire could be useful for these nutrients. As was reported in a study that proposed a new global food quality index ( 42 ) and taking into account the changes in dietary patterns resulting from globalisation, urbanisation and the greater availability of low-cost processed foods, it would be interesting to continue developing instruments that consider multiple aspects of dietary diversity, including more food groups, both healthy and unhealthy (ultraprocessed), and to evaluate their influence on the quality of diet and health. The CDA of this study allowed us to identify the amount of food consumed according to each food subgroup and food group, to identify whether the micronutrients consumed came from food or supplements, to identify populations at risk of deficient consumption of micronutrients and to establish policies or programmes that promote food production or nutrition education. This form is faster and less expensive to administer than 24HR, at 10 min v . 20 min. However, if necessary, it would be invaluable to develop an online application and generate food diversity software. One limitation of this study is that it was validated in a specific group of women of childbearing age who work in the same programme. Trained interviewers must have expertise in identifying the amount of food consumed, taking into account that the respondent must perform an extraction and condense several foods into one model. Conclusions The CDA validated in this study is useful to evaluate the population-level prevalence of the risk of deficiency in the usual intake of Ca, Fe, Zn, thiamine, riboflavin, niacin, pyridoxine, folates, vitamin B 12 , vitamin C and vitamin A. It was not useful to individually assess the prevalence of risk of deficiency in the usual intake of micronutrients, as the concordance analyses were weak and the ability to detect deficiencies in the diagnostic performance tests was only good for vitamin A, Ca, Fe, folates and pyridoxine. It is necessary to apply two CDA on non-consecutive days and distribute them throughout the week to adjust them in the PC-SIDE software. Although in the intramethod analysis (CDA), no significant differences were found in any micronutrients, when the prevalences between the 24HR method and CDA were compared without statistical adjustment, there were statistically significant differences in almost all micronutrients. A great variety of questionnaires, such as the one we have considered in this work, are useful instruments but are not meant to replace other instruments, such as 24HR recalls, that capture daily food consumption. Together with the appropriate statistical methodologies, 24HR recalls still provide the most precise assessment of usual intake distributions and the prevalence of inadequacy. Therefore, national-level interventions such as food fortification should still rely on the more precise individual-level, replicated, 24HR recalls. Acknowledgements Acknowledgements: The authors thank the Buen Comienzo programme, the interviewers, the data takers, and everyone who participated and made the completion of this research possible. Financial support: This work was funded by the Strategy for Sustainability of the Research Group on Food and Human Nutrition (Grupo de Investigación en Alimentación y Nutrición Humana, GIANH) of the University of Antioquia and by the Directorate of the School of Nutrition and Dietetics of the University of Antioquia, which provides economic support to projects of the Master of Science in Food and Human Nutrition (MCANH). The funding entities had no role in the design, analysis or writing of this article. Authorship: All authors of this study contributed equally to the processes of conceptualisation, analysis, research, methodology, project management, validation, writing, review and editing. Ethics of human subject participation: This study was conducted according to the guidelines laid down in the Declaration of Helsinki ( 1 ) and was classified with minimal risk according to Resolution 8430 of 1993 ( 2 ) . All procedures involving research study participants were approved by the ethics committee of the University Research Headquarters (SIU) of the University of Antioquia. Written informed consent was obtained from all participants. For confidentiality purposes, the questionnaires and personal data were coded and were only known by the principal investigators. At the end of the research, the general results were reported to the institutions and the individual results to the participating women. Conflict of interest: There are no conflicts of interest. Supplementary material For supplementary material accompanying this paper visit https://doi.org/10.1017/S1368980022000854. S1368980022000854sup001.docx (262.5KB, docx) click here to view supplementary material References 1. Morán L, Rivera A, González M et al. (2015) History of the valuation methods and applications of food consumption. Rev Esp Nutr Comunitaria 21, 17–23. [ Google Scholar ] 2. Colombian Institute of Family Welfare (2006) National Survey of the Nutritional Situation in Colombia 2005. Bogotá: ICBF. [ Google Scholar ] 3. Ministry of Social Protection of Colombia (2019) National Survey of the Nutritional Situation in Colombia 2015. Bogotá: ICBF. [ Google Scholar ] 4. Colombian Institute of Family Welfare (2017) National Study of the Food and Nutrition Situation of Indigenous Peoples in Colombia. Bogotá: ICBF. [ Google Scholar ] 5. MANÁ & University of Antioquia (2019) Food and Nutrition Profile of Antioquia 2019. Medellín: MANÁ. [ Google Scholar ] 6. Hoyos G (2015) Food and Nutritional Security Profile of Medellin and Its Suburbs 2015. Medellín: Municipal Government of Medellín. [ Google Scholar ] 7. Nutrition Institute of Central America and Panama (2006) Manual of Dietary Assessment Instruments. Guatemala: INCAP. [ Google Scholar ] 8. Zacarías I (1997) Dietary Evaluation Methods. Production and Management of Food Chemical Composition Data in Nutrition. Santiago: FAO. [ Google Scholar ] 9. Kennedy G, Ballard T & Dop M (2013) A Guide to Measuring Dietary Diversity at the Individual and Household Level. Rome: FAO. [ Google Scholar ] 10. Kennedy E, Ohls J, Carlson S et al. (1995) The Healthy Eating Index: design and applications. J Am Diet Assoc 95, 1103–1108. [ DOI ] [ PubMed ] [ Google Scholar ] 11. Patterson RE, Haines PS & Popkin B (1994) Diet Quality Index: capturing a multidimensional behavior. J Am Diet Assoc 94, 57–64. [ DOI ] [ PubMed ] [ Google Scholar ] 12. Huijbregts P, Feskens E, Räsänen L et al. (1997) Dietary pattern and 20 year mortality in elderly men in Finland, Italy, and the Netherlands: longitudinal cohort study. Br Med J 315, 25–30. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Trichopoulou A, Kouris-Blazos A, Wahlqvist ML et al. (1995) Diet and overall survival in elderly people. Br Med J 311, 1457–1460. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Gil Á, Martinez E & Olza J (2015) Indicators for the evaluation of diet quality. Rev Esp Nutr Comunitaria 21, 127–143. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Arimond M, Wiesmann D, Becquey E et al. (2010) Simple food group diversity indicators predict micronutrient adequacy of women’s diets in 5 diverse, resource-poor settings. J Nutr 140, 2059–2069. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Prevel Y, Becquey E & Arimond M (2010) Food group diversity indicators derived from qualitative list-based questionnaire misreported some foods compared to same indicators derived from quantitative 24-hour recall in Urban Burkina Faso. J Nutr 140, S2086–S2093. [ DOI ] [ PubMed ] [ Google Scholar ] 17. Food and Agriculture Organization (2015) Moving Forward: on Choosing a Standard Operational Indicator of Women’s Dietary Diversity. Rome: FAO. [ Google Scholar ] 18. Municipal Government of Medellín (2021) Buen Comienzo programme. https://www.medellin.edu.co/buen-comienzo/ (accessed January 2022). 19. Rothman KJ, Gallacher JEJ & Hatch EE (2013) Why representativeness should be avoided. Int J Epidemiol 42, 1012–1014. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Ministry of Health and Social Protection of Colombia (2016) Resolution 2465 of 2016. Whereby anthropometric indicators, reference patterns and cut-off points are adopted for the anthropometric classification of the nutritional status of children and adolescents under 18 years of age, adults from 18 to 64 years of age and adult pregnant women and other provisions are issued. Colombia. 21. Ferrari AM (2013) Intake estimation by means of a 24-hour reminder. Diaeta 31, 20–25. [ Google Scholar ] 22. Instituto of Medicine (2000) Minimizing Potential Errors in Assessing Group and Individual Intakes. Dietray Reference Intakes Applications in Dietary Assessment. Washington DC: National Academy Press. [ Google Scholar ] 23. Manjarrés LM (2008) Reliable data collected method about food intake in population studies. Persp Nut Hum 9, 155–163. [ Google Scholar ] 24. Cadavid M, Restrepo L, Rivillas J et al. (2006) Concordance between direct weigh of food portions consumed by children 5 to 9 years old, using tridimensional geometrical figures and interviewer memorization weight’s technique. Persp Nut Hum 15, 31–43. [ Google Scholar ] 25. Manjarrés L, Hernandez J & Cárdenas D (2015) Dietary Intake Screening Program (EVINDI, for its name in Spanish) v5. Medellín: University of Antioquia. [ Google Scholar ] 26. Quintero D, Alzate M & Moreno S (1990) Food Composition Table. Medellín: Nutritional Care Center. [ Google Scholar ] 27. Quintero D & Escobar L (2001) Food Composition Table. 2nd ed. Medellín: Nutritional Care Center. [ Google Scholar ] 28. Herrán O, Bautista L & Quintero D (2003) Table of Food Composition Consumed in Bucaramanga. Bucaramanga: Epidemiology Research Center, Industrial University of Santander. [ Google Scholar ] 29. Colombian Institute of Family Welfare (2005) Table of Colombian Food Composition. Bogotá: ICBF. [ Google Scholar ] 30. Colombian Institute of Family Welfare (2015) Table of Colombian Food Composition, 2nd ed. Bogotá: ICBF. [ Google Scholar ] 31. Food and Agriculture Organization (2020) International Network of Food Data System (INFOODS). http://www.fao.org/infoods/infoods/tables-and-databases/latin-america/en/ (accessed April 2021). 32. United States Department of Agriculture (2020) Food Composition Databases. https://fdc.nal.usda.gov/ (accessed April 2021). 33. Nusser S, Carriquiry A, Dodd K et al. (1996) A semiparametric transformation approach to estimating usual daily intake distributions. J Am Stat Assoc 91 1440–1149. [ Google Scholar ] 34. Ministry of Health and Social Protection of Colombia (2016) Resolution number 3803 of 2016. Whereby the RIEN Energy and Nutrient Intake Recommendations for the Colombian population are established and other provisions are issued. https://www.minsalud.gov.co/Normatividad_Nuevo/Resolución3803de2016.pdf (accessed April 2021). 35. Nusser SM, Carriquiry AL, Dodd KW et al. (1996) A User’s Guide to C-SIDE (Software for Intake Distribution Estimation). Version 1.0. Dietary Assessment Research Series Report 8. Iowa: University Statistical Laboratory. [ Google Scholar ] 36. Correa N & Manjarrés LM (2019) Compilation of Information on Nutrients in Food (RINAs). Medellín: University of Antioquia. [ Google Scholar ] 37. Navarro C, González E, Schmidt J et al. (2016) Profile of nutritional status of children and adolescents from Bogotá, Colombia. The FUPRECOL study. Nutr Hosp 33 915–922. [ Google Scholar ] 38. Zhang Q, Chen X, Liu Z et al. (2017) Diet diversity and nutritional status among adults in southwest China. PLOS ONE 12 1–9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. World Health Organization (2009) Indicators for Assessing Infant and Young Child Feeding Practices. Conclusions of the Consensus Meeting Held November 6–8, 2007 in Washington, DC, USA Part 1. Definitions. Geneva: WHO. [ Google Scholar ] 40. Abris G, Kim N-H, Provido S et al. (2018) Dietary diversity and nutritional adequacy among married Filipino immigrant women: the Filipino Women’s Diet and Health Study (FiLWHEL). BMC Public Health 18, 1–10. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Kennedy G, Fanou-Fogny N, Seghieri C et al. (2010) Food groups associated with a composite measure of probability of adequate intake of 11 micronutrients in the diets of women in Urban Mali. J Nutr 140, S2070–S2078. [ DOI ] [ PubMed ] [ Google Scholar ] 42. Ratner R, Hernández P, Martel J et al. (2017) A proposed new index of global food quality. Rev Chil Nutr 44, 33–38. [ Google Scholar ] 43. Watson PF & Petrie A (2010) Method agreement analysis: a review of correct methodology. Theriogenology 73, 1167–1179. [ DOI ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials For supplementary material accompanying this paper visit https://doi.org/10.1017/S1368980022000854. S1368980022000854sup001.docx (262.5KB, docx) click here to view supplementary material Articles from Public Health Nutrition are provided here courtesy of Cambridge University Press ACTIONS View on publisher site PDF (333.1 KB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top