![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
3 Nutrition and Consumer Protection Division, Food and Agriculture Organization, Rome, Italy 00153; 4 Division of Human Nutrition, Wageningen University, Wageningen, The Netherlands; 5 Food and Nutrition Research Institute, Metro Manilla, Philippines; and 6 Department of Agricultural Economics, University of Florence, Florence, Italy 50144
* To whom correspondence should be addressed. E-mail: s_g_kennedy{at}yahoo.com.
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
|
Introduction |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
Evidence from dietary intake research in the Philippines shows that the diets of a large percentage of young children are deficient in iron, vitamin A, and calcium. Intakes of vitamin C, niacin, riboflavin, and thiamin were found to be adequate. Average energy intake of preschool age children was also below recommended levels (2). Moving from a monotonous diet to one containing a more diverse range of foods has been shown to increase intake of energy as well as micronutrients in developing countries (3–7). Intake of a diverse variety of foods has been a recommendation for achieving adequate nutrient intake and the recommendation appears in the dietary guidelines of many countries. The nutritional guidelines for the Philippines include a number of recommendations on dietary diversity; 2 recommendations specify daily intake; 1) eat a variety of foods every day and 2) consume milk, milk products, and other calcium rich foods such as small fish and dark green leafy vegetables every day (8). Other recommendations encourage greater consumption of certain food groups but do not specify how often these should be consumed (fish, lean meat, poultry, dried beans, vegetables, fruits, and root crops). The precise number of foods or food groups that one should strive to consume over any given period is not commonly mentioned in most dietary guidelines. Japan advises consumption of 30 different food items per day (9) and the US advocates consumption of a variety of nutrient-dense foods and beverages within and among 5 basic food groups, with an item from each food group consumed daily [the 5 USDA food groups are: cereals, vegetables, fruit, dairy, and protein source foods (meat, fish, poultry, eggs, nuts, beans)] (10).
Despite many national nutritional guidelines recommending consumption of a variety of foods to meet nutritional needs, including those in the Philippines, the question remains how to operationalize this message for use as an indicator in the public health setting. The use of dietary diversity as an indicator of adequate nutrient intake remains under evaluation, particularly in developing countries. In those settings where the importance of dietary diversity to adequate nutrient intake has been assessed, researchers have used different food group classification systems, as well as diverse reference periods, cut-off points, and age groups (11). There is need for a set of comparable validation studies using the same methodology for creating a dietary diversity score (DDS)7 to predict adequate micronutrient intake.
The purpose of this study is to validate dietary diversity as an indicator of micronutrient adequacy in the diet of Filipino children 24–71 mo of age and to quantify the appropriate DDS cut-off point for use as an indicator of inadequate micronutrient intake. The results of this study will aid in the development and promotion of rapid assessment tools for measuring diversity of the diet and further understanding of the utility of a measure of dietary diversity as part of a set of indicators used to monitor food and nutrition security.
|
Materials and Methods |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
The food intake data of the preschool children were collected by individual 24-h food recall. The mother or caregiver was the respondent. The interviews included a detailed description of the foods eaten, the cooking method, and brand names (e.g. for milk consumed or other processed snack foods). The amount consumed by the child was estimated by the respondent, expressed in terms of cups, spoons, matchbox pieces, and other common household utensils. The respondents were shown visual aids to assist them in accurately reporting food intake. For mixed recipes, the respondent was asked how the food was prepared and how much of the visible components (e.g. pieces of meat, vegetables, etc.) were eaten by the child.
To compute nutrient values, the cooked weight was converted to raw weight using the Filipino Food and Nutrition Research Institute's Individual Dietary Evaluation Software. The software also contains a library of food composition values in their raw form. Nutrient values for energy, protein, fat, calcium, iron, vitamin A, vitamin C, thiamin, riboflavin, and niacin are from the 1997 food composition tables of the Philippines (13) and from food labels, particularly for iron and vitamin A, for fortified foods. For this study, nutrient values for vitamin B-6, vitamin B-12, folate, zinc, and phytate were obtained from the World Food Dietary Assessment System, version 2.0 (14). Nutrient retention values, from the USDA Table of Nutrient Retention Factors, Release 5 (2003), were added to account for nutrient losses during cooking process (15).
The data were cleaned for the purposes of this study. The average per capita daily energy requirement (kcal/d) for children 12–47 mo of age in the Philippines was calculated using the Population Energy Requirements software (16). The average per capita energy requirement was estimated at 4707 kJ/d (1125 kcal/d). The 5th and 95th percentiles, corresponding to intakes below 1607 kJ/d (384 kcal/d) and above 6632 kJ/d (1585 kcal/d) were discarded, leaving a total of 2805 records used in the analysis.
For the analysis using anthropometric data, only records with complete information on age, gender, weight, and height were included. WHO fixed exclusion ranges were used as criteria for cleaning outlying anthropometric Z scores (17).
DDS. DDS were calculated for each child using a set of 10 food groups (cereals and tubers; meat, poultry and fish; dairy; eggs; pulses and nuts; vitamin A-rich fruits and vegetables; other fruit; other vegetables; oils and fats; and other). The choice of the 10 food groups was based on the outcome of discussions held during a workshop on validation methods for dietary diversity held in Rome, Italy in October 2004. The decision was based on previous experience and testing of the usefulness of different food groupings (5) and is reflected in a set of basic guidelines for validating DDS in non-breast-feeding children 24–83 mo of age (18) and also in validation guidelines for children 0–24 mo of age (19). The food group "other," consisting of sugar, non-juice or dairy beverages, and condiments and spices, was used in descriptive statistics but was not used for tests of correlation, because this group does not contribute substantially to micronutrient intake. The majority of the analysis presented is based on the 9 food groups, excluding the "other" category.
DDS were calculated by summing the number of unique food groups consumed by the child in the 24-h period. An all inclusive DDS was calculated without a minimum intake for the food group. A second DDS was calculated applying a 10-g minimum intake for all food groups (DDS 10g) except fats and oils.
Nutrient bioavailability. Bioavailability adjustments were made for calcium, iron, and zinc. The purpose of making the bioavailability adjustments was to derive estimates of absorbed calcium, absorbed iron, and absorbed zinc to more accurately reflect concurrence between dietary intake and requirements. Bioavailability factors for calcium were 25% for roots, tubers, and legumes; 45% for fruits and vegetables; 5% for high oxalate vegetables (amaranth, cassava root and leaves, and spinach); and 32% for all other foods, based on Weaver et al. (20).
Bioavailability factors for iron were estimated at 6% for plant foods and 11% for animal source foods, based on a synthesis of sources, including FAO/WHO and Tseng et al. (21,22).
Bioavailability factors for zinc were calculated based on the phytate to zinc molar ratio. A ratio of 
18 was considered to have 30% bioavailability, whereas for a phytate to zinc ratio >18, a bioavailability factor of 22% was used based on calculations derived from Hotz and Brown (23).
Estimated average nutrient requirements and probability of adequate intake. The estimated average requirements (EAR) were used to assess the probability of adequate nutrient intake (PA). The EAR approach has been recommended as an improvement over using recommended nutrient intakes (RNI) for nutrient assessment of groups (24) as it allows for calculation of the probability that the individual's intake is adequate given the requirement distribution. The assumptions of the probability approach are that: 1) the requirement and intakes are independent; 2) the mean and variance of the requirement is known; and 3) the shape of the requirements distribution is known or can be assumed (25). The Institute of Medicine (IOM) report on applications of dietary reference intakes indicates that for all nutrients except energy, intakes and requirements are independent (24). The mean, variance, and distribution of requirements are known or calculated and assumed normal for all nutrients, with the exception noted in the IOM document of iron, where the distribution of requirements is skewed (24).
PA was calculated by the equation PA = PROBNORM [(estimated child intake – EAR)/SD], where PROBNORM is the statistical function that calculates the probability that a child's intake is above the EAR. The mean probability of adequate micronutrient intake (MPA) for each child is the average of the PA for the 11 micronutrients in the analysis. The mean PA and mean MPA were then calculated for the entire sample. The probability approach to assess adequacy of intake has been used in recent studies with a similar aim (5,26) and is also now part of the World Food Dietary Assessment System, version 2.0 (14). More information about the application of the probability approach can be found in the IOM report on applications in dietary assessment (24).
EAR for micronutrients. To derive EAR based on international requirements set by the United Nations, the EAR was back calculated from FAO/WHO RNI (Table 1). The RNI is defined as EAR+2SDEAR (21). The CV used to perform the calculations was based on IOM recommendations, set at 10% for all nutrients except 15% for niacin, 20% for vitamin A, and 25% for zinc (27–29).
|
Because iron requirements are not normally distributed, calculation for iron requirement and probability of adequate intake were derived from IOM iron requirements (29). Table 1-5 in that document was used as the basis for constructing a matrix for probability of adequate iron intake for children in age ranges 12–47 mo and 48–107 mo. We converted data in that table from 18% bioavailability to 10% bioavailability, which is more realistic of a high phytate, primarily vegetable-based diet (21) as typically consumed by children in the Philippines (Supplemental Table 1).
Statistical analysis. Statistical analysis was performed using SPSS version 11.5. PA and MPA were calculated separately for children 12–47 and 48–71 mo using respective EAR. Pearson's correlations were run by age group to verify the linear association for MPA and individual PA for each micronutrient. Linear regression models have been estimated separately for DDS and DDS 10g. DDS was evaluated for sensitivity and specificity using MPA as the gold standard. Sensitivity and specificity analysis were performed to quantify the accuracy of DDS to correctly classify children with high MPA values and then to determine the DDS cut-off point that maximized sensitivity and specificity. Two MPA cut-off values (0.50 and 0.75) were used in the analysis to categorize the children with low or high nutrient intakes.
|
Results |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
|
|
|
|
|
|
|
|
Discussion |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
In a study of school-aged children in Kenya, the mean DDS was 5.18 (based on 7 food groups) and mean MPA was 70% (5). In Kenya, the highest probability of inadequate intake for individual nutrients was zinc, vitamin B-12, calcium, vitamin E, and vitamin A. The results from our study were similar, with calcium, folate, and zinc having the lowest PA.
There are 2 similar validation studies on children of roughly the same age group in developing countries, 1 from South Africa and another from Mali. These studies used recommended dietary allowance instead of EAR to validate adequate intake of micronutrients and calculated nutrient adequacy ratios and a mean nutrient adequacy ratio (MAR) for each child. The study in South Africa found a mean DDS of 3.58 (based on the same 9 food groups used in this study) with a mean MAR of 50% (6). The nutrients with the lowest adequacy ratios were iron, calcium, and zinc. In Mali, the mean DDS was 5.8 (based on 8 food groups), with a mean MAR of 0.77 (4). The nutrients with the lowest nutrient adequacy ratio were riboflavin, calcium, vitamin A, and vitamin C.
The low intake of thiamin and riboflavin in this study was somewhat surprising, because these nutrients are present in most staple foods. Low PA of these nutrients in this study also differed from the results in Kenya (5). Rice has the lowest amount of thiamin and riboflavin per 100 g compared with wheat and maize, with maize being the staple food in Kenya, whereas the Filipino diet is based on rice. The practice of milling rice into highly polished white kernels removes an additional large percentage of thiamin. Highly milled polished rice contains roughly 0.06 mg thiamin/100 g, or only 12% of the EAR for a young child. Another explanation for the low intakes of thiamin and riboflavin in Filipino children comes from low milk consumption, particularly in children over the age of 1 y (1).
Our Pearson's correlation (0.36) between DDS and MPA was significant. The studies in Kenya, South Africa, and Mali also found significant correlations between DDS and nutrient intake: 0.39 (Mali), 0.32 (Kenya), and 0.64 (South Africa). In this study, using DDS 10g improved the correlation with MPA to 0.44, indicating that the performance of dietary diversity as an indicator of adequate micronutrient intake is improved when a minimum intake for each food group can be assessed. This finding has important implications for field use of the indicator, as collecting information on quantities of food consumed is more time consuming than simply recording the number of food groups consumed.
In our study, 2 nutrients, calcium and vitamin B-12, were not significantly correlated with DDS. Vitamin B-12 is found only in animal source foods, particularly liver, dairy products, and eggs. The best sources of calcium are dairy products, some legumes, green leafy vegetables, and small fish species, particularly if the bones are consumed. Dairy, eggs, and legumes were the least consumed food groups in the study population and lack of these groups could explain the poor correlation with DDS. Green leaves and fish were more commonly consumed, although the portion size consumed tended to be small. The lack of consumption of any foods from the dairy, egg, or legume group is the more likely explanation of poor correlation, as small portion sizes were common for most food groups except cereals/tubers.
A final aim of the study was to determine cut-off points for DDS, which can be used to classify children who are at greater risk of inadequate micronutrient intake. Similar to the Kenya study and using the 50th percentile of MPA, our results (not shown) found the best cut-off point to maximize sensitivity and specificity is a DDS of 5. However, the 50th percentile of our population corresponded to a mean MPA of 0.31, which may not be considered a sufficiently high enough cut-off to achieve an adequate improvement in population micronutrient intake. The results in Figure 2 test the sensitivity and specificity cut-off points using MPA of 0.50 and 0.75, a methodology previously applied by Hatloy et al. using MAR (4). Using MPA of 0.50 and 0.75, the best cut-off point for maximizing both sensitivity and specificity is between DDS of 5 and 6. Determining a fixed cut-off point where children can be defined as having greater or less risk of inadequate micronutrient intake has potential application in both immediate population nutritional assessment and continued monitoring of improvement in micronutrient intake. The ultimate decision as to which is the most appropriate MPA to use to define the DDS cut-off point, as well as whether it is more desirable to maximize sensitivity or specificity or find the point that optimizes both, will depend on the desired use of the DDS indicator. For example, if the goal of the indicator is to maximize identification of at-risk children, one would aim to maximize sensitivity; however, this would reduce specificity, thereby including more children who are not truly at risk in the target group. One potential use of the DDS is as an international indicator of risk of inadequate micronutrient intake. To realize this objective, additional validation studies using the same methodology for datasets from different geographic and cultural settings should be replicated.
One limitation of the study is that only 1 24-h recall was available per child; therefore, it was not possible to correct for within-person variation of intake. Not accounting for this variation could affect the MPA as well as perhaps the DDS cut-off point. Future studies should test the use of the indicator after adjusting for within-person variation in intake.
The aim of this study was to determine how well a simple score of food groups can be used to predict adequate micronutrient intake. The results have shown that DDS is correlated with MPA and also that DDS is a significant determinant of MPA. Using the more rigorous measure of DDS 10g did improve the correlation and regression model. Additionally, energy intake had a strong influence on MPA.
Current methods used to assess micronutrient deficiencies primarily rely on biochemical diagnostic tests of blood or urine, which, although considered the gold standard, are often difficult, time consuming, and expensive to collect and analyze, and are thus not generally widely used in community settings for monitoring and evaluation of nutrition improvement programs. There is a need to develop convenient, cost efficient indicators that can measure changes in the micronutrient status of vulnerable populations. This paper demonstrates that a simple count of food groups can be used to predict the probability of adequate micronutrient intake in young non-breast-feeding Filipino children. Indices that include additional information such as quantities of food consumed or total energy intake should enhance the performance of the indicator. The decision about the level of detail to incorporate into a survey will depend on the time available for data collection, overall study budget, and purpose or objective for which the indicator will be used.
|
FOOTNOTES |
|---|
2 Supplemental Tables 1 and 2 are available with the online posting of this paper at jn.nutrition.org.
7 Abbreviations used: DDS, dietary diversity score; DDS 10g, dietary diversity score with 10-g minimum consumption; EAR, estimated average requirement; IOM, Institute of Medicine; MAR, mean adequacy ratio; MPA, mean probability of adequate micronutrient intake; PA, probability of adequate micronutrient intake; RNI, recommended nutrient intake.
Manuscript received 6 May 2006. Initial review completed 19 June 2006. Revision accepted 20 November 2006.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTIntroductionMaterials and MethodsResultsDiscussionLITERATURE CITED |
|---|
1. Pedro MRA, Cerdena CM, Molano WL, Consatntino A, Perlas LA, Palafox FF, Patalan L, Chavez M, Madriaga J, et al. The current state of nutrition in the Philippines: children in focus [monograph on the Internet]. Food and Nutrition Research Institute; 2005 [cited 2006 March 29]. Available from: http://www.fnri.dost.gov.ph/.
2. Pedro MRA, Benavides RC, Barba CVC. Dietary changes and their health implications in the Philippines in the double burden of malnutrition: case studies from six developing countries. Rome: Food and Agriculture Organization. In press 2006.
3. Hoddinott J, Yohannes Y. Dietary diversity as a food security indicator. Washington: International Food Policy Research Institute, Food Consumption and Nutrition Division; 2002. Discussion Paper No. 136.
4. Hatloy A, Torheim LE, Oshaug A. Food variety: a good indicator of nutritional adequacy of the diet? A case study from an urban area in Mali, West Africa. Eur J Clin Nutr. 1998;52:891–8.[Medline]
5. Ruel M, Graham J, Murphy S, Allen L. Validating simple indicators of dietary diversity and animal source food intake that accurately reflect nutrient adequacy in developing countries. Report submitted to GL-CRSP; 2004.
6. Steyn N, Nel J, Nantel G, Kennedy G, Labradarios D. Food variety and dietary diversity scores; are they good indicators of dietary adequacy? Public Health Nutr. 2006;9:644–50.[Medline]
7. Onyango A, Koski K, Tucker K. Food diversity versus breastfeeding choice in determining anthropometric status in rural Kenyan toddlers. Int J Epidemiol. 1998;27:484–9.
8. FNRI (Food and Nutrition Research Institute). Nutritional guidelines for Filipinos. Revised ed. Manila (Philippines): FNRI; 2000.
9. Truswell S. Evolution of dietary recommendations, goals and guidelines. Am J Clin Nutr. 1987;45:1060–72.
10. USDA. Dietary guidelines for Americans 2005 [monograph on the Internet]. Washington: USDA [cited March 1, 2006]. Available from: http://www.healthierus.gov/dietaryguidelines/.
11. Ruel M. Operationalizing dietary diversity: a review of measurement issues and research priorities. J Nutr. 2003;133:S3911–26.
12. Villavieja G, Cerdena C, Molano W, Lana R, Boquecosa J, Raymundo B, Nones C, Abaya H, Palafox E, et al. Fourth National Nutrition Survey, Philippines, 1993. Part A. Food consumption survey. Philipp J Nutr. 1997;44:1–33.[Medline]
13. Food and Nutrition Research Institute. The Philippine food composition tables. Revised ed. Manila (Philippines): FNRI; 1997.
14. User's Guide to the World Food Dietary Assessment System. [cited 2005 June 28]. Accessed as a Word document from the International Network of Food Data Systems (INFOODS). Available from: http://www.fao.org/infoods/.
15. USDA, Agricultural Research Service. USDA National Nutrient Database for Standard Reference, Release 18. USDA Nutrient Retention Factors, Release 5. 2003 Nutrient Data Laboratory [database on the Internet]. Washington: USDA [cited 2006 July 17]. Available from: http://www.nal.usda.gov/fnic/foodcomp
16. FAO. Human energy requirements. Rome: FAO; 2004. FAO Food and Nutrition Technical Report Series 1. Report of a Joint FAO/WHO/UNU expert consultation.
17. WHO. Physical status: the use and interpretation of anthropometry. Geneva: WHO; 1995. WHO Technical Report Series 854.
18. Kennedy G, Nantel G. Basic guidelines for validating dietary diversity score as an indicator of adequate micronutrient intake; 2006 [cited 2006 April 15]. Available from: ftp://ftp.fao.org/ag/agn/nutrition/dds_validation.pdf
19. Arimond M, Cohen R, Dewey K, Ruel M. Developing and validating simple indicators of complementary food intake and nutrient density for infants and young children in developing countries: protocol for data analysis. Washington: IFPRI/UC Davis; 2005. Report submitted to the Food and Nutrition Technical Assistance Project, Academy for Educational Development.
20. Weaver CM, Proulx WR, Heaney R. Choices for achieving adequate dietary calcium with a vegetarian diet. Am J Clin Nutr. 1999;70:S543–548.
21. FAO/WHO. Human and vitamin mineral requirements. Rome: FAO; 2002. Report of a Joint FAO/WHO expert consultation.
22. Tseng M, Chakraborty H, Robinson D, Mendez M, Kohlmeier L. Adjustments of iron intake for dietary enhancers and inhibitors in population studies: bioavailable iron in rural and urban residing Russian women and children. J Nutr. 1997;127:1456–68.
23. Hotz C, Brown K, editors. Assessment of the risk of zinc deficiency in populations and options for its control. International Zinc Nutrition Consultative Group. Technical Document No. 1. Food Nutr Bull. 2004;25 Suppl 2:S94–203.
24. Institute of Medicine. Dietary reference intakes: applications in dietary assessment. Washington: National Academy Press; 2000a.
25. Carriquiry AL. Assessing the prevalence of nutrient inadequacy. Public Health Nutr. 1999;2:23–33.[Medline]
26. Foote J, Murphy S, Wilkens L, Basiotis P, Carlson A. Dietary Variety increases the probability of nutrient adequacy among adults. J Nutr. 2004;134:1779–85.
27. Institute of Medicine. Dietary reference intakes for vitamin c, vitamin e, selenium, and carotenoids. A report of the Panel on Dietary Antioxidants and Related Compounds and the Standing Committee on the Scientific Evaluation of Dietary Reference Intakes. Washington: National Academy Press; 2000b.
28. Institute of Medicine. Dietary reference intakes for thiamin, riboflavin, niacin, vitamin b6, folate, vitamin b12, pantothenic acid, biotin and choline. A report of the Standing Committee on the Scientific Evaluation of Dietary Reference Intakes. Washington: National Academy Press; 2000c.
29. 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: National Academy Press; 2001.
30. UK Department of Health. Dietary reference values for food energy and nutrients for the United Kingdom. London: Department of Health; 1991.
This article has been cited by other articles:
|
|
 |
|
  M. M. Moursi, M. Arimond, K. G. Dewey, S. Treche, M. T. Ruel, and F. Delpeuch Dietary Diversity Is a Good Predictor of the Micronutrient Density of the Diet of 6- to 23-Month-Old Children in Madagascar J. Nutr., December 1, 2008; 138(12): 2448 - 2453. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M Moursi, Y. Martin-Prevel, S. Eymard-Duvernay, G. Capon, S. Treche, B. Maire, and F. Delpeuch Assessment of child feeding practices using a summary index: stability over time and association with child growth in urban Madagascar Am. J. Clinical Nutrition, May 1, 2008; 87(5): 1472 - 1479. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
