«Minimum Dietary Diversity for Women A Guide to Measurement Minimum Dietary Diversity for Women A Guide to Measurement Published by the Food and ...»
Institute of Medicine. 2005. Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids (Macronutrients). Washington, DC, The National Academies Press.
Lee, S.E., Talegawkar, S.A., Merialdi, M. & Caulfield, L.E. 2013. Dietary intakes of women during pregnancy in low- and middle-income countries. Public Health Nutr. 16(8): 1340–53. Doi: 10.1017/ S1368980012004417.
Liu, R.H. 2013. Health-promoting components of fruits and vegetables in the diet. Adv Nutr. 4(3):
384S–92S. doi: 10.3945/an.112.003517.
Malik, V.S., Pan, A., Willett, W.C. & Hu, F.B. 2013. Sugar-sweetened beverages and weight gain in
children and adults: a systematic review and meta-analysis. Am J Clin Nutr. 98(4): 1084–102. Doi:
Martin-Prével, Y., Allemand, P., Wiesmann, D., Arimond, M., Ballard, T., Deitchler, M., Dop, M.C., Kennedy, G., Lee, W.T. & Mousi, M. 2015. Moving forward on choosing a standard operational indicator of women’s dietary diversity. Rome: FAO.
National Research Council. 2006. Dietary Reference Intakes: The Essential Guide to Nutrient Requirements. Washington, DC, The National Academies Press.
Ros, E. 2015. Nuts and CVD. Br J Nutr. 113(Suppl 2): S111–20. Doi: 10.1017/S0007114514003924.
Ruel, MT. 2003. Operationalizing dietary diversity: a review of measurement issues and research priorities. J Nutr. 133(11 Suppl 2): 3911S–3926S.
Rumpold, B.A. & Schlüter, O.K. 2013. Nutritional composition and safety aspects of edible insects.
Mol Nutr Food Res. 57(5): 802–23. Doi: 10.1002/mnfr.201200735.
Torheim, L.E. & Arimond, M. 2013. Diet quality, micronutrient intakes and economic vulnerability of women. In V.R. Preedy, L.A. Hunter & V.B. Patel, eds. Diet Quality: An Evidence-Based Approach, Vol. I, pp. 105–115. New York, Springer.
Turati, F., Rossi, M., Pelucchi, C., Levi, F. & La Vecchia. C. 2015. Fruit and vegetables and cancer risk: a review of southern European studies. Br J Nutr. 113(Suppl 2): S102–10. Doi: 10.1017/ S0007114515000148.
WFP. 2008. Food consumption analysis: Calculation and use of the food consumption score in food security analysis. Technical Guidance Sheet. Rome, WFP.
WHO. 2008. Indicators for assessing infant and young child feeding practices. Part I: Definitions.
WHO/FAO. 2004. Vitamin and mineral requirements in human nutrition: Report of a joint FAO/WHO expert consultation, Bangkok, Thailand, 21–30 September 1998. 2nd edition. Geneva, WHO.
WHO/FAO. 2006. Guidelines on food fortification with micronutrients. Geneva, WHO.
Xi, B., Huang, Y., Reilly, K.H., Li, S., Zheng, R., Barrio-Lopez, M.T., Martinez-Gonzalez, M.A. & Zhou, D. 2015. Sugar-sweetened beverages and risk of hypertension and CVD: a dose-response metaanalysis. Br J Nutr. 113(5): 709–17. Doi: 10.1017/S0007114514004383.
Appendix 1. Sampling and design issues specific to measurement of Minimum Dietary Diversity for Women of Reproductive Age There are numerous sampling, sample size and survey design decisions that depend on the objectives and context for data collection, but a general discussion of these issues is beyond the scope of this manual.
However, there are a few decisions specific to measurement of food group diversity for women of reproductive age (WRA), including selection of respondent(s) within the household, sampling of days of the week, sampling of “unusual” days (e.g. feasts) and issues related to seasonality.
Selection of respondent(s) within the household There are two options for selection of respondent women within the household: selection of all ageeligible women or random selection of one age-eligible woman. Note that all age-eligible women include those considered to be living in the household, even if not present at the time the survey team visits.
Age For both options, first screen women on age and select those who have reached their 15th birthday but who have not yet reached their 50th birthday. In cases where exact birth dates or ages are unknown, local calendars are sometimes used to help establish estimated ages26.
Number of respondents The decision on whether to include all women in the household or to randomly select one will depend on the broader sample design decisions.
Random selection of one of several women within a household requires use of appropriate sampling weights during analysis to avoid under-representation of women who live in larger households with multiple WRA. Calculation of sampling weights in turn requires information on the number of ageeligible women in the household.
Selection of all age-eligible women results in non-independent observations and this too must be handled appropriately during data analysis.
Either choice is valid, so long as analysis methods and inferences account correctly for this choice.
Sampling of days of the week Every effort should be made to collect data on all days of the week. People may eat differently on different days of the week; this is part of the overall diet and part of the picture of diet quality at population level. If days of the week are represented with equal frequency in the data set, eating patterns will also be properly represented. If it is not possible to collect data on all days of the week, for example, for legal or cultural reasons associated with work on the Sabbath, it is still important to have data collection take place on the other 6 days.
See, for example: FAO. 2008. Guidelines for Estimating the Month and Year of Birth of Young Children (available at 26 http://www.ifad.org/hfs/docs/guidelines.pdf). This document describes the use of local calendars.
Sampling “unusual” days In general, there is no need to avoid using feast days, weddings or other celebration days as the day recalled by the respondent, for the same reason noted above in relation to sampling all days of the week. It is fine if some individuals in the sample have consumed more than usual, for one reason or another, on the day recalled. This is part of normal variation in intakes.
However, if a large proportion of a community has participated in a special feast or celebration, it is better not to visit (sample) that community the following day, as the recall day would be unusual for the entire community.
Ramadan presents a specific problem because of its duration and because eating patterns may be different for many or all members of the community as compared with all other times of year. Except in the context of surveys that are rolling or that sample the entire year, it is better to avoid fielding food group diversity surveys during Ramadan. If it is necessary to field during Ramadan, this should be considered during interpretation of results.
While certain days of the week and/or celebrations may entail increased and more varied intakes, intakes and variety may be lower than usual when people are ill. However, there is no need to avoid sampling or using data from days when respondents report that they had low appetite or illness on the day recalled. This too is part of normal variation in intakes within a population on any given day.
In summary, unusual intakes at the individual level are not a problem and should not be treated differently during data collection or analysis. However, when there is reason to believe intakes for an entire community or a large segment of the community would be highly unusual, it is better to avoid surveying at that time.
Seasonality Diet patterns in many contexts vary with season. For example, mango season may strongly affect the proportion of women reporting consumption of vitamin A-rich fruits and thus may affect the proportion reaching the threshold of five or more food groups. Other seasonal foods may have less impact; for example, in some settings, types of green leafy vegetables vary with season, but one type or another is consumed year-round.
It is also possible for food group diversity to increase during lean/hunger seasons, when foraged foods may be consumed. These foods may add diversity, and even micronutrients, but in the context of inadequate caloric intakes. In this situation, an increase in diversity cannot be viewed in isolation.
Survey designers should consider seasonality when fielding and when interpreting results from food group diversity surveys. In particular, avoid direct comparisons between surveys conducted during different seasons, if it is apparent that seasonality could affect diversity in the context27. Similarly, avoid direct comparisons between surveys conducted in different geographic areas experiencing different seasons – for example, do not compare results from the hunger season in one zone to the post-harvest season in another, even if they occur in the same month.
Ideally, food group diversity indicators should not be compared unless there are rolling surveys covering all seasons or the indicators are generated from surveys carried out in the same season.
When data and capacity allow (e.g. in research contexts), it is also possible to adjust for seasonality using the survey date and geographic positioning system data.
This challenge is not unique to data collection for the Minimum Dietary Diversity for Women of Reproductive Age (MDD-W) indicator; seasonality is an issue for many food security, health and nutrition indicators.
The exception would be in research or other contexts where survey designers aim to capture and describe 27 seasonal variation, as the topic of study.
Appendix 2. Guidance on assigning individual foods to food groups for Minimum Dietary Diversity for Women of Reproductive Age28 This appendix provides extensive examples for each of the rows on the Minimum Dietary Diversity for Women of Reproductive Age (MDD-W) model questionnaire in Section 3 of the guide.
Fourteen groups (rows) on the questionnaire are used to construct the ten MDD-W food groups.
Several of the MDD-W food groups are further divided on the questionnaire (e.g. meat and poultry are separated from fish and seafood). These 14 rows (A–N) are followed by 6 optional categories (rows O–T on the questionnaire) and two required final categories (row U, “Condiments and seasonings”, and row V, “Other beverages and foods”).
Examples are provided for rows A–V on the model questionnaire. At the end of this appendix, following the examples for row V, there is a table providing guidance on typical classification challenges.
Note that the row order differs in the alternative list-based questionnaire in Appendix 3. This is because when using the list-based method, it is necessary to consider how the order of the foods listed might influence responses to avoid double-counting of certain foods. This is not an issue in an open recall. See Appendix 3 for further explanation.
This appendix can be used during adaptation of the questionnaire.
When listing example items in each row of the questionnaire, use local names for foods. This is especially important for staple foods and other groups where the source ingredient typically undergoes processing (commercially or in the home). For example, rather than listing “wheat” on the questionnaire, list local food names, such as bread, chapatti, noodle, pasta, roti, seitan and/or wheat tortilla. Similarly, in the pulses group, be sure to list hummus, tofu and/or other locally consumed processed products made from pulses. In other food groups, food and ingredient names may be fine (e.g. most fruits and vegetables).
This section is adapted from: WHO. 2010. Indicators for assessing infant and young child feeding practices Part 2:
28 Measurement. WHO: Geneva.
A. Foods made from grains Include products and foods derived from cereal crops. Any staple dishes or products like breads (e.g.
bagels, rolls, chapatti, roti, tortillas), porridge (ugali, nsima/nshima, posho, sadza, mealies, dalia, muesli, papilla, grain fufu) and noodles (pasta, soba, spaghetti, vermicelli) made from the grains listed below, and from flours of these grains, should be included in this category.
Sweet biscuits and cakes are not included and are classified with “Sweets” (category “S” below).
B. White roots and tubers and plantains Include non-coloured items mainly providing carbohydrate. This group includes all non-grain-based starchy staples. Any staple dishes/casseroles and pastes made from roots, tubers and plantains should also be included in this category.
C. Pulses (beans, peas and lentils) This group includes members of the plant family Fabaceae (alternate name Leguminoseae), including beans, peas and lentils. The seeds are harvested at maturity and dried and used as food or processed into a variety of food products. This group does not include the same plants harvested green/ immature and eaten fresh in the pod – these are included in the “Other vegetables” group (group “M”). It also does not include groundnut (peanut), because while groundnut is in the Fabaceae family, both its high fat content and most common culinary uses are different from other legumes and similar to those of tree nuts. Groundnut is included in the “Nuts and seeds” group (group “D”).
The pulses group includes mature seeds (beans), sprouted pulses and processed/prepared products, such as hummus, tofu, tempeh, soy milk, soy cheese, texturized vegetable protein and other soy products and products of any of the pulses listed in the table.
D. Nuts and seeds This group comprises mostly tree nuts but also includes groundnut (peanut) and may include certain seeds when consumed in substantial quantities. Defining “seeds” for inclusion in this category is challenging; see discussion below the table. In many cases, seeds should be included in category “U” (Condiments and seasonings).
This group also includes nut and seed “butters”, such as pounded groundnut/peanut butter, cashew butter or sesame butter (tahini) when consumed in substantial amounts and not merely added to flavour mixed dishes.
Note that oils extracted from nuts and seeds are not included in this group; they are included in “Other oils and fats” (category “Q”).