Dietary species richness as a measure of food biodiversity and nutritional quality of diets

Carl Lachata,1,2, Jessica E. Raneria,b,1, Katherine Walker Smitha, Patrick Kolsterena, Patrick Van Dammec,d, Kaat Verzelenc, Daniela Penafielc,e, Wouter Vanhovec, Gina Kennedyb, Danny Hunterb, Francis Oduor Odhiambob, Gervais Ntandou-Bouzitoub, Bernard De Baetsf, Disna Ratnasekerag, Hoang The Kyh, Roseline Remansa,b, and Céline Termoteb aDepartment of Food Safety and Food Quality, Faculty of Bioscience Engineering, Ghent University, 9000 Ghent, Belgium; bHealthy Diets from Sustainable Food Systems Initiative, Bioversity International, 00057 Maccarese (Rome), Italy; cLaboratory of Tropical and Subtropical Agronomy and Ethnobotany, Faculty of Bioscience Engineering, Ghent University, 9000 Ghent, Belgium; dDepartment of Crop Sciences and Agroforestry, Faculty of Tropical AgriSciences, Czech University of Life Sciences Prague, 165 21 Prague 6, Suchdol, Czech Republic; eRural Research Center, Faculty of Life Sciences, Nutrition, Escuela Superior Politecnica del Litoral, Guayaquil, 090608 Ecuador; fKERMIT, Department of Mathematical Modeling, Statistics, and Bioinformatics, Faculty of Bioscience Engineering, Ghent University, 9000 Ghent, Belgium; gDepartment of Agricultural Biology, Faculty of Agriculture, University of Ruhuna, 81100 Matara, Sri Lanka; and hHealthBridge Foundation of Canada, 10000 Hanoi, Vietnam

Edited by David Tilman, University of Minnesota, St. Paul, MN, and approved November 9, 2017 (received for review June 6, 2017)

Biodiversity is key for human and environmental health. Available (10). Wild food diversity, obtained in or around agricultural fields dietary and ecological indicators are not designed to assess the or extracted from forests and other natural landscapes, is an ad- intricate relationship between food biodiversity and diet quality. ditional source of resilience in the food system, in particular during We applied biodiversity indicators to dietary intake data from and the lean season (9). Adequate management and use of biodiversity assessed associations with diet quality of women and young can help to restore ecosystems and address micronutrient defi- children. Data from 24-hour diet recalls (55% in the wet season) of ciencies in vulnerable populations (11). n = 6,226 participants (34% women) in rural areas from seven low- Surprisingly, the world’s wild and agricultural biodiversity hot and middle-income countries were analyzed. Mean adequacies of spots often coincide with low-income areas with high poverty vitamin A, vitamin C, folate, calcium, iron, and zinc and diet diversity levels, ecosystem degradation, and malnutrition (12, 13). Re- score (DDS) were used to assess diet quality. Associations of bio- duced biodiversity of both wild and agricultural species can have diversity indicators with nutrient adequacy were quantified using detrimental effects for diet quality and environmental sustain- multilevel models, receiver operating characteristic curves, and test sensitivity and specificity. A total of 234 different species were con- ability by reducing availability and access to nutritious, seasonal sumed, of which <30% were consumed in more than one country. foods and loss of ecosystem functions (14). Sustainable man- — Nine species were consumed in all countries and provided, on average, agement of food biodiversity the diversity of plants, animals, 61% of total energy intake and a significant contribution of micro- nutrients in the wet season. Compared with Simpson’s index of di- Significance versity and functional diversity, species richness (SR) showed stronger associations and better diagnostic properties with micronutrient ad- Current research linking biodiversity and human diets has used equacy. For every additional species consumed, dietary nutrient ade- metrics without justification from a nutritional point of view. quacy increased by 0.03 (P < 0.001). Diets with higher nutrient Diet species richness, or a count of the number of different adequacy were mostly obtained when both SR and DDS were max- species consumed per day, assesses both nutritional adequacy imal. Adding SR to the minimum cutoff for minimum diet diversity and food biodiversity of diets for women and children in rural improved the ability to detect diets with higher micronutrient ade- areas. The positive association of food species richness with quacy in women but not in children. Dietary SR is recommended as dietary quality was observed in both the wet and the dry season. the most appropriate measure of food biodiversity in diets. Food biodiversity contributes to diet quality in vulnerable pop- ulations in areas with high biodiversity. Reporting the number of sustainable diets | diet quality | malnutrition | biodiversity | species consumed during dietary assessment provides a unique food biodiversity opportunity to cut across two critical dimensions of sustainable development—human and environmental health—and comple- ood systems are a key driver of biodiversity loss worldwide ments existing indicators for healthy and sustainable diets. F(1). Globally, key drivers of food system transformations in- clude climate change, population growth, economic development, Author contributions: C.L., J.E.R., R.R., and C.T. designed research; C.L. and J.E.R. performed research; J.E.R., G.K., and H.T.K. contributed data from Vietnam; D.P. contributed data from urbanization, globalization, and production system intensification Ecuador; K.V., D.H., and D.R. contributed data from Sri Lanka; G.N.-B. contributed data from and homogenization (2–4). As a result, human diets that used to Benin; P.V.D. contributed data from Democratic Republic of Congo, Ecuador, and Cameroon; be composed of a wide variety of plants and animals have gradually W.V. contributed data from Cameroon; F.O.O. contributed data from Kenya; C.T. contributed shifted to a diet composed of mostly processed foods and com- data from Democratic Republic of Congo, Benin, Kenya, and Cameroon; J.E.R., P.K., P.V.D., and R.R. contributed new reagents/analytic tools; C.L., K.W.S., P.K., B.D.B., and R.R. analyzed prising a limited number of species (5). While an estimated 300,000 data; and C.L., J.E.R., K.W.S., P.K., P.V.D., K.V., D.P., W.V., G.K., D.H., F.O.O., G.N.-B., B.D.B., D.R., edible plant species are available to humans, more than half of the H.T.K., R.R., and C.T. wrote the paper. SCIENCES global energy need is currently met by only four crops: rice, potatoes, The authors declare no conflict of interest. wheat, and maize (6). This article is a PNAS Direct Submission. APPLIED BIOLOGICAL Low-quality diets are the leading risk factor for ill health This open access article is distributed under Creative Commons Attribution-NonCommercial- worldwide (7) and are determined by socioeconomic and political NoDerivatives License 4.0 (CC BY-NC-ND). factors including income, education, social cohesion, gender em- Data deposition: Anonymized individual-level data and protocols for each country are powerment, and inequality (8). The diversity of species used in publicly available (https://dataverse.harvard.edu/dataverse/DietarySpeciesRichness). agricultural and livelihood systems is essential for human nutrition 1C.L. and J.E.R. contributed equally to this work. and sustainable food systems (9). Agricultural biodiversity con- 2To whom correspondence should be addressed. Email: [email protected]. SCIENCE tributes to farm resilience, particularly in the face of shocks such This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. SUSTAINABILITY as climate change, disease outbreaks, and market price fluctuations 1073/pnas.1709194115/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1709194115 PNAS | January 2, 2018 | vol. 115 | no. 1 | 127–132 and other organisms used for food, both cultivated and from the species richness (SR), a count of the number of species consumed by each individual; wild—is essential for sustainable food systems (15). Simpson’s index of diversity (D), which represents the number of different species While hunger, food security, and sustainability are addressed in consumed and how evenly the amounts consumed of these different species are the Sustainable Development Goals (SDGs), the current indicators distributed based on quantity consumed; and the functional diversity (FD), as the used for SDGs 2 and 15 capture nutritional status, sustainable total branch length of a functional dendrogram. FD reflects the diversity in nu- trient composition of species consumed by each individual (25). The three metrics management of terrestrial ecosystems, and agricultural sustainability represent different aspects of diversity [i.e., SR, evenness and richness combined dimensions separately and do not consider diet quality or food (D), FD]. Here, these are regarded as food biodiversity indicators. biodiversity loss (16). Evidence within the context of sustainable diets SR was calculated as a count of the number of different species (plants or is particularly limited when it comes to human diet and biodiversity animals) consumed by an individual. D was calculated using the ineq package in (17). Research linking biodiversity, agricultural production diversity, Stata, taking into account the weight of species consumed (grams) in relation to and human diets has used multiple metrics without validation from a the total weight of all species consumed per individual per day. Similar to previous dietary point of view (18). Existing diet diversity indicators such as studies (26), FD was calculated with the nutritional composition of the species the diet diversity score (DDS) or food variety score are used as consumed [i.e., content of the vitamin A (retinol activity equivalent μg/100 g), proxies for dietary quality and measure the diversity of unique food vitamin C (mg/100 g), folate (μg/100 g), calcium (mg/100), iron (mg/100 g), and groups and food items consumed, respectively (19). Neither of these zinc (mg/100 g) regarded as traits using the picante and ade4 library in R]. indicators specifically captures the biological contribution of diverse Foods and drinks were identified at the species level following best-practice plant and animal species to human diets. Luckett et al. (20) applied guidelines (27) using local taxonomical references and technical expertise. Some the nutritional functional diversity score to diets, but within the studies, especially those with an ethnobotanical focus, had thorough documen- tation of species verified through a local herbarium, botanical museum, or context of measuring the contribution of different food source out- botanist or biologist. Only data from Vietnam and Sri Lanka identified variety or lets to diet diversity and without validating the nutritional adequacy breed level where possible. Species-level data were utilized to calculate the of the measure. We argue that indicators to monitor progress in biodiversity indicators. When it was not possible to identify the specific species, achieving healthy and environmentally sustainable diets must inte- the identified genus was followed by “sp.” Given specific disambiguation chal- grate diet quality and biodiversity (21). lenges (28), all bananas were recorded as Musa sp. Disambiguation of scientific The present study aimed to recommend a cross-cutting in- names was performed using published databases (The Plant List, www.theplantlist. dicator that measures food biodiversity in human diets and helps org/; Species 2000 & ITIS Catalogue of Life, www.catalogueoflife.org/col). guide interventions toward human and environmental health Taxonomical information was not available if the food consumed was a mixed simultaneously. We applied three ecological biodiversity indica- preparation where the ingredients and quantities of ingredients were un- < tors to dietary intake data of women and children in seven low- known (five recipes accounting for 0.04% of the total energy consumed in the and middle-income countries and evaluated how these indicators sample). When the only food consumed in a food group was not identifiable at species level, at least one species was counted for that specific food group to were associated with nutrient adequacy. Associations between calculate the SR. Because the study assessed the level of food biodiversity in the food biodiversity, diet diversity, and nutrient adequacy as three diet, intake of breast milk was not considered in the calculation of biodiversity complementary dimensions of diet quality were examined. Finally, indicators. The consumption of different parts of particular plant or animal we assessed the use of a cutoff for minimal food biodiversity to species was counted once, with no minimum quantity. No minimum quantity identify diets with higher nutrient adequacy and compared it with consumed was applied to include a species in the biodiversity indicators. the existing cutoff for minimum diet diversity. Nutritional Indicators. Food-composition data for vitamin A, vitamin C, calcium, Methods folate, iron, and zinc were mostly sourced from national food-composition Data Sources. Existing data were first mapped using a systematic literature tables (SI Appendix,TableS1). In the case that variety-level information was review (22). The search syntax was applied to 10 databases: (i) Agricola, available, this was used to calculate nutrient intake. When food-composition (ii) Agris, (iii) Bioline International, (iv) EMBASE, (v) IngentaConnect, data were missing, best-matching values were obtained from similar settings, (vi) Web of Knowledge, (vii) Medline (through PubMed), (viii) Science Direct, countries, or foods. Adjustments for bioavailability were considered as per (ix) Cochrane Library of Systematic Reviews, and (x) Worldcat. A combination previous protocols for calcium (29), iron (30), and zinc (31). of “(food OR diet OR nutrition) AND biodiversity,” was used as syntax and As a measure of nutritional quality, the mean adequacy ratio (MAR) was tailored to the database. A detailed syntax was reported elsewhere (22). The calculated as the arithmetic mean of the quantity of a nutrient consumed per its database of papers was further searched for studies on dietary quality using requirement for each individual on a daily basis. Individual nutrient adequacy “energy,”“energy intake,”“micronutrient,”“dietary diversity,” and “food ratios (NAR) were capped at 1, so nutrients with high levels of consumption diversity” as keywords. Only studies on food biodiversity and human nutri- could not compensate for those with lower levels when calculating MAR. Higher tion were considered, while those on animal nutrition, biofuels, simulations, values of MAR correspond to a higher adherence of the diet to nutritional re- microbiology, and genetically modified organisms were excluded. Researchers quirements for the micronutrients included in the MAR. Estimated Average who reported species consumed in diets were contacted to identify relevant Requirements were used from FAO (30), the Institute of Medicine (32), and the data (23). We selected datasets that (i)usedaquantitativeandcomparable European Food Safety Authority (33). data collection method, (ii) were concerned with identifying foods and drinks The DDS for women was a count of the total number of food groups to at least the species level, and (iii) assessed dietary intake of either women or consumed from a list of 10: (i) grains, white roots and tubers, and plantains; children or both. Researchers that supervised these studies collaborated for the (ii) pulses; (iii) nuts and seeds; (iv) dairy; (v) meat, poultry, and fish; (vi) eggs; present analysis. All studies used an interviewer-administered quantitative (vii) dark-green leafy vegetables; (viii) other vitamin A-rich vegetables and 24-h recall and considered wild and agricultural sources of food. Because not all fruits; (ix) other vegetables; and (x) other fruits (34). For children, a seven- studies had a repeated recall, only the first day was included. Study charac- food-group classification was used, including the following: (i) grains, white teristics and data collection methods are summarized in SI Appendix,TableS1. roots and tubers, and plantains; (ii) legumes, nuts, and seeds; (iii) dairy; The Food and Agricultural Organization of the Food and Agriculture Organi- (iv) meat, poultry, and fish; (v) eggs; (vi) vitamin A-rich fruits and vegetables; zation (FAO) and the International Institute for Applied Systems Analysis and (vii) other fruits and vegetables (35). As recommended (34, 35), a 15-g (IIASA) global agro-ecological zone and FAO–World Bank farming systems minimum quantity consumed was considered as a cutoff for species inclusion classification were used to describe the sample area (24). in the DDS for women but not for children. The Minimum Dietary Diversity Food intake data during the wet season were obtained from rural areas in (MDD) was used as a cutoff for higher nutrient adequacy and refers to a Benin, Cameroon, the Democratic Republic of Congo, Ecuador, Kenya, Sri minimum of five and four food groups for women and children, respectively. Lanka, and Vietnam. Data from the dry season were also available in Vietnam, Except in Sri Lanka where the protocol was exempted from clearance, all Kenya, and Benin (SI Appendix, Table S1). All data were collected between studies were approved by an ethics committee. The present analysis was July 2009 and April 2015, and samples were representative of the village approved by the Ethics Committee of Ghent University (NR B670201422403). population. Anonymized individual-level data and protocols are available (https://dataverse.harvard.edu/dataverse/DietarySpeciesRichness). Data Analysis. Because every sample was considered equally representative, the overall summary statistics were calculated averages for women and Food Biodiversity Indicators. We calculated three types of diversity metrics based children separately, per country and across countries. We compared mean on all species (plant, livestock, and fish) consumed over the 24-h recall period: MAR, DDS, and food biodiversity indicators between women and children or

128 | www.pnas.org/cgi/doi/10.1073/pnas.1709194115 Lachat et al. Table 1. MAR, DDS, and food biodiversity indicators in women and children by country and season

Benin Cameroon Congo Ecuador Kenya Sri Lanka Vietnam All (n = 2,188 (n = 2,439 (n = 125 (n = 462 (n = 258 (n = 790 women, (n = 36 women, (n = 642 women, women, 4,038 children) children) women) women) 790 children) 20 children) 664 children) children)

Indicators Wet Dry Wet Wet Wet Wet Dry Wet Wet Dry Wet Dry

MAR Women ———0.64 ± 0.15 0.65 ± 0.12 0.62 ± 0.16 0.71 ± 0.15 0.53 ± 0.16 0.69 ± 0.13 0.67 ± 0.14 0.63 ± 0.06 0.69 ± 0.03 Children 0.44 ± 0.24 0.46 ± 0.24 0.61 ± 0.16 ——0.57 ± 0.20 0.64 ± 0.19 0.68 ± 0.18 0.70 ± 0.22 0.69 ± 0.16 0.60 ± 0.10 0.60 ± 0.12 DDS Women ———2.89 ± 1.27 5.28 ± 1.21 3.93 ± 1.13 4.13 ± 1.11 4.00 ± 1.07 4.67 ± 1.24 4.07 ± 1.20 4.16 ± 0.90 4.11 ± 0.05 Children 4.05 ± 0.97 4.08 ± 1.01 3.62 ± 0.97 ——3.85 ± 0.95 3.95 ± 0.99 4.45 ± 1.36 4.14 ± 1.14 4.12 ± 1.12 4.00 ± 0.33 4.06 ± 0.08 SR Women ———9.64 ± 3.57 16.39 ± 3.09 8.20 ± 2.02 8.51 ± 1.84 8.08 ± 2.95 9.04 ± 3.33 8.16 ± 2.88 10.27 ± 3.48 8.34 ± 0.25 Children 9.00 ± 3.20 9.21 ± 3.33 7.90 ± 2.45 ——8.88 ± 2.43 9.42 ± 2.51 8.80 ± 4.09 6.40 ± 2.74 6.29 ± 2.57 8.22 ± 1.11 8.31 ± 1.75 D Women ———0.84 ± 0.11 0.94 ± 0.02 0.81 ± 0.12 0.84 ± 0.10 0.87 ± 0.05 0.87 ± 0.06 0.83 ± 0.05 0.87 ± 0.05 0.84 ± 0.01 Children 0.79 ± 0.13 0.80 ± 0.13 0.85 ± 0.07 ——0.87 ± 0.46 0.88 ± 0.08 0.90 ± 0.33 0.88 ± 0.08 0.86 ± 0.06 0.86 ± 0.04 0.85 ± 0.04 FD Women ———0.74 ± 0.2 0.95 ± 0.26 0.45 ± 0.11 0.52 ± 0.10 0.53 ± 0.21 0.43 ± 0.12 0.42 ± 0.11 0.62 ± 0.22 0.48 ± 0.07 Children 0.57 ± 0.15 0.50 ± 0.12 0.56 ± 0.16 ——0.45 ± 0.13 0.53 ± 0.11 0.57 ± 0.27 0.30 ± 0.11 0.31 ± 0.10 0.50 ± 0.12 0.45 ± 0.12 Unique 11 17 8 17 36 9 13 14 51 65 143 98 species*

Congo, Democratic Republic of Congo. *No. of different species that were only consumed in one country. Means and SDs are tabulated. seasons using t test with averages per country. Means ± SDs are reported. To NAR were also comparable, except for vitamin A, which was par- quantify the association between the measures of food biodiversity con- ticularly higher in the wet season (SI Appendix,TableS3). The sumed and the micronutrient adequacy of diets, a random-effects model quantity of staple food consumption was only marginally higher in was used accounting for different associations per country. The model in- cluded season as a fixed-effects variable and used an unstructured covariance the dry season. The average quantity of pulses, dark-green leafy matrix. To ensure comparable estimates, food biodiversity indicators were vegetables, and vitamin A-rich vegetables consumed was notably expressed as z scores in these models and standardized coefficients were used. higher (>15 g) in the dry season than in the wet season (SI Appendix, Because intake of species is potentially associated with total dietary energy Table S2). The average number of species consumed per food group intake, we also performed the analysis after adjusting the food biodiversity was comparable in the dry and wet season (SI Appendix,TableS4). indicators for energy intake using the residual method (36). Associations be- A total of 234 different species were consumed by participants tween variables were visualized using locally weighted regression curves. We and mean SR was lower in children than in women (8.24 ± 1.17 vs. used heat maps with the mean MAR per DDS for the different food biodiversity ± < indicators to assess the associations of food biodiversity across food group di- 10.19 3.52; P 0.001). The species consumed per country are versity and micronutrient adequacy. We compared test characteristics of the included as SI Appendix,TablesS5andS6. An average of 1.73 ± food biodiversity indicators and DDS to identify diets with higher nutrient ad- 0.94 species were consumed per food group. Less than one-third of equacy using receiver operating characteristic (ROC) curves. Finally, we assessed if adding a component of food biodiversity to the MDD cutoff would increase the ability to define higher nutrient adequacy compared with MDD. To account for nutritional differences and the contribution of species between food groups, the product of DDS × SR was used for this purpose. An MAR >50% was considered a threshold for minimal nutrient adequacy. Similar to the validation of MDD (37), a minimal sensitivity (Se) and specificity (Sp) of 60% was used to determine a DDS × SR cutoff. Test diagnostics properties were compared using ROC curves and test Se and Sp. Stata 14 (StataCorp) was used for data analysis. Results Dietary intake data were obtained for n = 3,449 (55%) and n = 2,777 participants during the wet and dry season, respectively. Women (n = 2,188; 34%) were mainly of childbearing age (mean age: 31.0 ± 11.7 y). Apart from n = 32 Kenyan children, all children (n = 4,038) were between 6 and 24 months old. On average, 94% of the energy intake was identified at the species level. Items that were not identified at the species level were sweets, water, salt, and bicarbonate and food items with missing species information at data collection. For processed food specif-

ically, only five foods accounting for 0.04% of total energy from SCIENCES food were not identified. Of foods included in the DDS, >93% were identified at the species level. Those foods that were not APPLIED BIOLOGICAL assigned to a food group of the DDS were consumed in small quantities (∼5g/d)(SI Appendix,TableS2). MAR was comparable for children and women (0.61 ± 0.09 vs. 0.63 ± 0.06; P = 0.85; Table 1). Diets were particularly in- adequate with regard to iron (SI Appendix, Table S3). MAR,

DDS, and food biodiversity indicators were comparable across SCIENCE

seasons when only the countries with data on the two seasons Fig. 1. Association of biodiversity indicators with MAR for 6,226 women SUSTAINABILITY were used (P = 0.90, P = 0.93, and P = 0.51, respectively; Table 1). and children in seven countries (wet and dry season combined).

Lachat et al. PNAS | January 2, 2018 | vol. 115 | no. 1 | 129 Table 2. Association between biodiversity measures and MAR Unstandardized Standardized

Biodiversity measures β SE β SE

SR 0.03 0.001 0.10 0.003 D 0.70 0.02 0.08 0.08 FD 0.52 0.02 0.06 0.002

Mixed-effects linear regression model with season (fixed effects) and country as random effects. All β coefficients are P < 0.001.

Fig. 3. Association of MAR with SR for 6,226 women and children in seven species were consumed in more than one country. Overall, 58% countries (wet and dry season combined). DR, Democratic Republic. (n = 143) and 40% (n = 98) of species were consumed only in a single country in the wet and dry season, respectively. Of all spe- Because of its stronger and consistent associations, and sim- cies, 53% (n = 125) were consumed during both seasons and 40% = = plicity in application, we used SR for further evaluation as a food (n 93) and 7% (n 16) of the species were unique to the wet or biodiversity indicator. The dry season and being a woman were dry season, respectively. associated with an average increase of 0.03 and 0.01 (both P < In the wet season, nine species [Arachis hypogaea L., Bos taurus 0.001) in MAR per additional species consumed, respectively. For Linnaeus, 1758, Glycine max (L.) Merr, Manihot esculenta Crantz, vitamin A, vitamin C, folate, and calcium, nutrient adequacy in- Oryza sativa L., Solanum lycopersicum L., Solanum tuberosum L., creased by 0.07 and for iron and zinc by 0.02 and 0.05, respectively Sus scrofa Linnaeus, 1758, and Zea mays L.] were consumed in all (all P < 0.001) per additional species consumed. Adjusting the countries and provided, on average, 61%, 10%, 24%, 42%, 51%, models for total energy intake did not modify the findings. 65%, and 35% of the total energy, vitamin A, vitamin C, folic acid, The best cutoff to define a diet with higher nutrient adequacy iron, zinc, and calcium intakes, respectively. In the dry season, was an MAR of 50% (SI Appendix,TableS7). The area under the 19 species (including all species that were common to all countries curve for DDS, SR, and DDS × SR to define MAR ≥50% was in the wet season except for S. tuberosum L.)werecommontoall comparable for both women and children (Fig. 4). In brief, adding three countries with dry season data and provided 87%, 35%, 31%, SR to DDS considerably increased the ability to identify higher- 51%, 74%, 85%, and 45% of the energy, vitamin A, vitamin C, folic quality diets in women. Compared with the MDD, a cutoff of 24 acid, iron, zinc, and calcium intakes on a daily basis, respectively. DDS × SR increased Se by 39% in women, with an acceptable Sp All three food biodiversity indicators were positively associ- (SI Appendix, Tables S8 and S9). Although test Se of a DDS × SR ated with MAR (Fig. 1). Per increase in SR, D, or FD z score, cutoff was lower compared with MDD, overall acceptable Se and MAR increased, on average, by 0.03, 0.7, and 0.52, respectively Spestimateswereobtainedinchildren(Table3). (Table 2). The standardized coefficients indicate that SR has a slightly stronger association with MAR than D and FD. Compared Discussion with the other food biodiversity indicators, ROC analysis also indicated To our knowledge, no previous studies have applied common a slightly higher ability for SR to define higher nutrient-adequate measures of biodiversity to measure levels of food biodiversity in diets (Fig. 2). The positive association with nutrient adequacy was the diet. All three biodiversity indicators assessed food bio- consistent between the countries for SR (Fig. 3) and FD (SI Ap- diversity in the diet and were positively associated with micro- pendix,Fig.S1) but less apparent for Simpson’sindex(SI Appendix, nutrient adequacy. SR showed stronger and more consistent Fig. S2). MAR increased with both the SR and DDS (Fig. 4). The associations with diet quality indicators (MAR and DDS) than associations of MAR and DDS with Simpson’s index and FD, Simpson’s index of D index and FD. Given that SR can more however, were less consistent (SI Appendix, Figs. S3 and S4). easily be calculated in comparison with D and FD, we recom- mend Dietary Species Richness (DSR) as the most appropriate measure of food biodiversity in diets. Decision makers often struggle to reconcile environmental and food policies. DSR is a valuable tool in this regard, because it in- tegrates biodiversity, nutrition, and health aspects of food systems. The use of an indicator such as DSR offers an opportunity to capture both biodiversity and dietary quality with a single metric. The positive association found between DSR and MAR was consistent across countries, populations, and both seasons. The present findings demonstrate a wide species diversity consumed by rural populations in low- and middle-income countries. The majority of species consumed were unique to each study site, highlighting the importance of local food biodiversity to diets. Micronutrient adequacy and DSR were similar in both seasons despite seasonal changes in the local production system and in- creased food availability associated with the dry season. This was unexpected, because an earlier systematic literature review (38) reported considerable intra-annual variations in diet quality. How- ever, none of these reviewed studies considered the consumption of underutilized, wild, or semiwild foods. Communities are not entirely composed of subsistence farmers. It is possible that households have Fig. 2. ROC curves of standardized biodiversity indicators with micronutrient supplemented their diets with foods sourced from the market and adequacy in women and children. MAR50, diet with 50% mean adequacy of the wild to compensate for changes or decreases in local food vitamin A, vitamin C, folate, calcium, iron, and zinc; zFD, standardized FD; zD, production availability. Unfortunately, information on food sources standardized D; zSR, standardized SR. (own production, market, wild) was not included in the analysis.

130 | www.pnas.org/cgi/doi/10.1073/pnas.1709194115 Lachat et al. 8 1.00 MAR50 in women 0.90 1.00 0.80 6 0.70 0.75 0.60

4 0.50

0.40 0.50 Sensitivity 0.30 SRxDDS ROC area: 0.79 2

Women dietary diversity Women 0.20 0.25 SR ROC area: 0.74 0.10 Mean nutrient adequacy ratio 0.00 DDS ROC area: 0.79 0 0.00 0 5 10 15 20 25 0.00 0.25 0.50 0.75 1.00 Species richness in women 8 1.00 MAR50 in children 0.90 1.00 0.80 6 0.70 0.75 0.60

4 0.50

0.40 0.50 Sensitivity 0.30 SRxDDS ROC area: 0.64 Child dietary diversity 2 0.20 0.25 SR ROC area: 0.61 0.10 Mean nutrient adequacy ratio 0.00 DDS ROC area: 0.66 0 0.00 0 5 10 15 20 0.00 0.25 0.50 0.75 1.00 Species richness in children

Fig. 4. MAR against SR and DDS (Left) and ROC curves for SR × DDS, SR, and DDS (Right) for 6,226 women and children in seven countries. MAR50, diet with 50% mean adequacy of vitamin A, vitamin C, folate, calcium, iron, and zinc.

As a result, it is not possible to perform a more in-depth analysis of in women. The improvement in test diagnostic properties, however, the exact role markets or wild foods played in the diet. Lack of was small and not observed in children. seasonal variation in the diet may also be explained by differences in On the other hand, assessing DSR can be challenging, because it the local production systems in terms of primary crops, harvest was estimated that previous studies misidentified between 6% and periods, time to receive income after harvesting, and lean seasons. 10% of species (39). Guidelines were recently prepared to ade- It is recommended that studies further examine how DSR corre- quately record species during food-intake studies (15). Using an lates with diet quality in the lean and abundant seasons, rather open recall or species-level food list during MDD data collection than in different climatic seasons, and further also consider the would also enable DSR calculation. Cost-effective technologies sources of foods consumed to better understand how diet quality is and approaches, such a mobile apps, that enable MDD enumer- maintained across different climatic seasons. ators to identify and record species-level details of foods consumed DSR was more strongly associated with MAR in the dry season, can be helpful in population diet-quality surveillance surveys. suggesting that it may be easier to increase nutrient adequacy in the Foods not classified in a DDS food group for women essen- dry season. This may be attributed to the observed higher quanti- tially contained species consumed as condiments or spices, or that ties consumed of legumes, vitamin A-rich fruits and vegetables, and were consumed in small quantities. These foods were included in dark-green leafy vegetables in the dry season. The availability of the food biodiversity indicators (when the species could be identified) these foods is highly seasonal. Innovative processing and storage and in MAR calculations. The large number of foods consumed, but methods and the introduction of species and varieties that are not captured by the DDS, highlights the contribution of these bio- productive “out of season” may extend their availability in the wet diverse foods that are consumed in small serving sizes, but with likely season when smaller quantities were consumed. nutritional benefits. We used DSR, which captures both agricultural and wild food Identifying food species diversity in diets is a useful first step biodiversity. Our study therefore does not reveal the contribution toward sustainability assessment of diets. Adding additional es- of agricultural vs. wild biodiversity in the diet. Earlier research timates on the environmental impact or ecosystem services (40) of the species consumed (e.g., chicken vs. beef vs. pork) would showed that DDS is positively associated with farm production allow for better assessment and modeling of the sustainability diversity as well as with market access (18). The contribution of of the diet. Such assessment will improve assessment of the wild biodiversity to the dietary quality is less clear (9). Future environmental and natural resource impacts from agricultural intake assessments should record the source of each food item to production or from extraction from natural ecosystems (41). shed more light on the relative contribution of locally available As is the case with other studies (42), a limitation of the present agricultural and wild biodiversity to dietary quality. This is im- work is a lack of nutrient-composition data of some foods, species, portant because it has implications for biodiversity conservation and varieties consumed. The composition of various indigenous, management in which focus on agricultural diversity might be at the expense of wild biodiversity conservation and vice versa. Scholars have called for dietary indicators that consider mul- tiple dimensions to provide more comprehensive assessments of Table 3. Test classification properties of SR and DDS cutoffs for > diet quality (18). DDS is a common measure to assess diversity higher dietary quality (MAR 50%) SCIENCES and diet quality and is widely applied in population surveys. The SR and DDS cutoffs Sensitivity, % Specificity, %

joint use of DSR and DDS ensures that complementary di- APPLIED BIOLOGICAL mensions of diet quality and diversity are included during dietary Women ≥ assessments. We report a positive association between DSR and DDS 5* 42 96 × ≥ DDS. Diets with higher nutrient adequacy were observed when DDS SR 24 81 60 both DSR and diet diversity were maximal. The DSR thus captures Children ≥ † both the dimension of biodiversity as well as diet diversity. The DDS 4 84 35 DDS × SR ≥36 62 54 combined application of both DDS and DSR as a minimum cutoff SCIENCE

combining food biodiversity and food group diversity concepts *Minimum DDS for women. SUSTAINABILITY improved the ability to detect diets with higher nutrient adequacy †Minimum DDS for children.

Lachat et al. PNAS | January 2, 2018 | vol. 115 | no. 1 | 131 wild, neglected, or underutilized species was often not available deficiencies (11). Monitoring the contribution of species in the diet and was substituted with values from similar foods. It is expected enables identifying species with the greatest potential to improve that wider identification of species and varieties consumed will diets in different local contexts and provides additional granularity guide food-composition assessment toward nutritionally relevant to assess the importance of food diversity in ensuring diet quality. and currently undocumented species. We used a single 24-h recall Global datasets such as FAOSTAT identify general food items or per subject. Although this method is appropriate to estimate pop- food groups and do not facilitate valorization of the full range of ulation average intakes, it does not allow accounting for within- food biodiversity. In addition, international food security efforts person variability and estimation of usual dietary intake. have hence focused on the production of a handful of staple foods Finally, we used dietary intake data from rural areas of middle- (mostly cereals) to meet human energy needs (6). The present study income countries where locally produced food is the major con- provides evidence on the role of nonstaple foods to both energy and tributor to diets. Food systems in (peri)urban areas and high-income countries have a higher degree of complexity than in rural areas and micronutrient intakes in rural areas. Identifying foods consumed at middle-income countries. This complexity is mainly caused by the species level adds information that supports both conservation the consumption of processed foods that have often not been and sustainable food system initiatives. locally produced but have been obtained from retail outlets or urban markets. Nevertheless, in diets with higher contributions ACKNOWLEDGMENTS. The research was funded by the Agriculture for Nutrition and Health (A4NH) CGIAR Research Programme (CRP). The of processed foods, we expect all three biodiversity indicators following sources funded studies from which the data was used: Benin: to remain a valid measure of food biodiversity. Depending on Ministry of Foreign Affairs Finland (FoodAfrica project); Cameroon and processing and fortification practices, however, the strength of Ecuador: Flemish Interuniversity Council; Congo: Flemish Interuniversity the association between the food biodiversity indicators and Council, Leopold III fund for Nature Exploration and Conservation, and diet quality may differ from the present findings. Further as- Stichting Roeping; Sri Lanka: Global Environment facility, United Nations sessment of the validity and applicability of DSR in diets with a Environmental Programme, Food and Agriculture Organization, and Bio- higher contribution of foods obtained from urban markets or versity International; and Kenya and Vietnam: Humidtropics and A4NH CRPs. R.R. received a grant from Daniel and Nina Carasso Foundation for research of processed foods is warranted. on nutrition-sensitive landscapes. The funders of the study had no role in Nutrition-sensitive agricultural and ecosystem conservation in- study design, data collection, analysis or interpretation, or writing of the terventions, specifically those related to diversification, clearly have manuscript. The corresponding author had full access to all of the data in the an untapped potential to address global hunger and micronutrient study and had final responsibility for deciding to submit for publication.

1. Maxwell SL, Fuller RA, Brooks TM, Watson JE (2016) Biodiversity: The ravages of guns, 23. Smith K, et al. (2014) Scoping Study to Determine the Data Sources on Biodiversity in nets and bulldozers. Nature 536:143–145. Diet and Food Intake (Ghent University, Ghent, Belgium). 2. Tilman D, Clark M (2014) Global diets link environmental sustainability and human 24. FAO; International Institute for Applied Systems Analysis (2016) GAEZ Global agro- health. Nature 515:518–522. ecological zones. Available at www.fao.org/nr/gaez/en/. Accessed December 11, 2016. 3. Pingali P (2007) Westernization of Asian diets and the transformation of food sys- 25. Wood SA, et al. (2015) Functional traits in agriculture: Agrobiodiversity and ecosys- – tems: Implications for research and policy. Food Policy 32:281 298. tem services. Trends Ecol Evol 30:531–539. 4. Gomez MI, Ricketts KD (2013) Food value chain transformations in developing 26. Remans R, et al. (2011) Assessing nutritional diversity of cropping systems in African – countries: Selected hypotheses on nutritional implications. Food Policy 42:139 150. villages. PLoS One 6:e21235. 5. Drewnowski A, Popkin BM (1997) The nutrition transition: New trends in the global 27. Nesbitt M, McBurney RPH, Broin M, Beentje HJ (2010) Linking biodiversity, food and – diet. Nutr Rev 55:31 43. nutrition: The importance of plant identification and nomenclature. J Food Compost 6. FAO (2010) Second Report on the State of the World’s Plant Genetic Resources for Anal 23:486–498. Food and Agriculture (FAO, Rome). Available at www.fao.org/3/a-i4465e.pdf. 28. Crichton R, Vezina A, Van den Bergh I (2014) An online checklist of banana cultivars. 7. Abajobir A, et al. (2017) Global, regional, and national comparative risk assessment of 84 be- Acta Hortic 1114:13–18. havioural, environmental and occupational, and metabolicrisksorclustersofrisks,1990-2016: 29. Kennedy G, Nantel G (2006) Basic Guidelines for Validation of a Simple Dietary A systematic analysis for the Global Burden of Disease Study 2016. Lancet 390:1345–1422. Diversity Score as an Indicator of Dietary Nutrient Adequacy for Non-Breastfeeding 8. Darmon N, Drewnowski A (2015) Contribution of food prices and diet cost to socioeconomic disparities in diet quality and health: A systematic review and analysis. Nutr Rev 73:643–660. Children 2-6 Years (FAO, Rome). 9. Powell B, et al. (2015) Improving diets with wild and cultivated biodiversity from 30. FAO/WHO (2004) Human Vitamin and Mineral Requirements (FAO, Rome), 2nd Ed. across the landscape. Food Secur 7:535–554. 31. Brown KH, et al.; International Zinc Nutrition Consultative Group (IZiNCG) (2004) 10. Zimmerer KS (2015) Understanding agrobiodiversity and the rise of resilience: Analytic Assessment of the risk of zinc deficiency in populations and options for its control. category, conceptual boundary object or meta-level transition? Resilience 3:183–198. Food Nutr Bull 25(Suppl 2):S99–S203, IZiNCG Technical Document 1. 11. Ruel MT, Alderman H; Maternal and Child Nutrition Study Group (2013) Nutrition- 32. Institute of Medicine (2006) Dietary Reference Intakes: The Essential Guide to Nu- sensitive interventions and programmes: How can they help to accelerate progress in trient Requirements (Natl Acad Press, Washington, DC), p 560. Available at nap.edu/ improving maternal and child nutrition? Lancet 382:536–551. 11537. Accessed October 27, 2016. 12. Herrero M, et al. (2017) Farming and the geography of nutrient production for hu- 33. European Food Safety Authority (2014) Scientific opinion on dietary reference values man use: A transdisciplinary analysis. Lancet Planet Health 1:e33–e42. for zinc. EFSA J 12:3844. 13. Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity 34. FAO; FHI360 (2016) Minimum Dietary Diversity for Women: A Guide to Measurement hotspots for conservation priorities. Nature 403:853–858. (FAO, Rome). Available at www.fao.org/3/a-i5486e.pdf. Accessed December 4, 2016. 14. Myers SS, et al. (2013) Human health impacts of ecosystem alteration. Proc Natl Acad 35. WHO (2008) Indicators for assessing infant and young child feeding practices. Part I: Sci USA 110:18753–18760. Definitions. Conclusions of a consensus meeting held 6–8 November 2007, Wash- 15. FAO and Bioversity International (2017) Guidelines on Assessing Biodiverse Foods in ington DC, USA (WHO, Geneva). Available at apps.who.int/iris/bitstream/10665/43895/ Dietary Intake Surveys (FAO, Rome). Available at www.fao.org/3/a-i6717e.pdf. Ac- 1/9789241596664_eng.pdf. Accessed December 1, 2016. cessed June 20, 2017. 36. Willett WC, Howe GR, Kushi LH (1997) Adjustment for total energy intake in epide- 16. United Nations (2017) Report of the Inter-Agency and Expert Group on Sustainable miologic studies. Am J Clin Nutr 65(Suppl):1220S–1228S, discussion 1229S–1231S. Development Goal indicators. E/CN.3/2017/2. Available at https://unstats.un.org/unsd/ 37. Martin-Prével Y, et al. (2015) Moving Forward on Choosing a Standard Operational Indicator statcom/48th-session/documents/2017-2-IAEG-SDGs-E.pdf. Accessed July 14, 2017. of Women’s Dietary Diversity (FAO, Rome). Available at www.fao.org/3/a-i4942e.pdf. Ac- 17. Garnett T (2016) Plating up solutions. Science 353:1202–1204. cessed December 4, 2016. 18. Jones AD (2017) Critical review of the emerging research evidence on agricultural 38. Phalkey RK, Aranda-Jan C, Marx S, Höfle B, Sauerborn R (2015) Systematic review of biodiversity, diet diversity, and nutritional status in low- and middle-income coun- tries. Nutr Rev 75:769–782. current efforts to quantify the impacts of climate change on undernutrition. Proc Natl – 19. Steyn NP, Nel JH, Nantel G, Kennedy G, Labadarios D (2006) Food variety and dietary diversity Acad Sci USA 112:E4522 E4529. Ł scores in children: Are they good indicators of dietary adequacy? Public Health Nutr 9:644–650. 39. uczaj LJ (2010) Plant identification credibility in ethnobotany: A closer look at Polish 20. Luckett BG, DeClerck FA, Fanzo J, Mundorf AR, Rose D (2015) Application of the nutrition ethnographic studies. J Ethnobiol Ethnomed 6:36. functional diversity indicator to assess food system contributions to dietary diversity and 40. van der Werf HMG, Petit J (2002) Evaluation of the environmental impact of agri- sustainable diets of Malawian households. Public Health Nutr 18:2479–2487. culture at the farm level: A comparison and analysis of 12 indicator-based methods. 21. FAO (2015) The Second International Conference on Nutrition: Committing to a future Agric Ecosyst Environ 93:131–145. free of malnutrition. Available at www.fao.org/3/a-ml542e.pdf. Accessed January 24, 2017. 41. Sachs J, et al. (2010) Monitoring the world’s agriculture. Nature 466:558–560. 22. Penafiel D, Lachat C, Espinel R, Van Damme P, Kolsteren P (2011) A systematic review on the 42. Ruth Charrondière U, et al. (2013) FAO/INFOODS food composition database for contributions of edible plant and animal biodiversity to human diets. EcoHealth 8:381–399. biodiversity. Food Chem 140:408–412.

132 | www.pnas.org/cgi/doi/10.1073/pnas.1709194115 Lachat et al. Lachat et al. Dietary species richness as a measure of food biodiversity and nutritional quality of diets SI Appendix: Supplementary figures and tables

Fig. S1. Association of mean nutrient adequacy ratio with functional diversity score for 6,226 women and children in 7 countries (wet and dry season combined)

1

Fig. S2. Association of mean nutrient adequacy ratio with Simpson’s index of diversity for 6,226 women and children in 7 countries (wet and dry season combined)

2

Fig. S3. Mean nutrient adequacy ratio against Simpson’s diversity index and dietary diversity score (left) and receiver operating characteristic curves for Simpson’s index and diet diversity score (right) for 6,226 women and children in 7 countries ROC: Receiver operating characteristic; DDS: Diet diversity score; MAR50: Diet with 50% adequacy of vitamin A, vitamin C, folate, calcium, iron and zinc

3

Fig. S4. Mean nutrient adequacy ratio against Functional diversity score and dietary diversity score (left) and receiver operating characteristic curves for Functional diversity and diet diversity score (right) for 6,226 women and children in 7 countries ROC: Receiver operating characteristic; DDS: Diet diversity score; FD: Functional diversity score; MAR50: Diet with 50% adequacy of vitamin A, vitamin C, folate, calcium, iron and zinc

4 Table S1 Description of studies

Country Benin Cameroon DR Congo Ecuador Kenya Sri Lanka Vietnam

Location Mono department Northwest region Oriental province* Cotopaxi province Vihiga county Ratnapura district Mai Son district

Data collection Oct-Dec 2013 / May-Jul Jul–Aug 2013 Jul–Sep 2009 Mar 2011 Sep 2014 & Apr 2015 Jul-Sep 2013 Aug- Dec 2014 period 2014

Agro-ecological Tropics lowland, warm Tropics lowland, Tropics lowland, Tropics lowland, Tropics lowland, warm Tropics lowland, warm Tropics highland, cool zone warm warm warm

Farming system Root crop Tree crop Forest based Highland intensive Maize mixed Rice Highland intensive mixed mixed

Population per Dry: Children (n=1220, Wet: Children Wet: Women (n=462, Wet: women Dry: women (n=396, mean Wet: women (n=36, Dry: women (n=374, mean season mean age=14.5 months) (n=125, mean, age mean age= 37.1 year) (n=258, mean age age 29.7 year), Children mean age = 47.1 year) age =24.0 year), Children =18.0 months) = 8.2 year) n=397 (mean age =17.4 (n=390, mean age 15.6 Wet: Children (n=1219, months) Children (n=20, mean months,) mean age=14.5 months) age = 11.6 months) Wet: women (n=394, Wet: women (n=268, mean mean age 30.7 year), age =23.9 year), Children Children n=393 (mean age (n=274, mean age 12.9 = 19.1 months) months)

Species Botanist from the ICRAF data for Local flora (2) in Local herbarium East African Herbarium, of Herbarium specimens Local ethnobiologist and identification national herbarium (1), Cameroon and herbarium of (3) National Museums of collected and identified head of agriculture extension procedure an ichthyologist and a World Vegetable Yangambi DR Congo Kenya (4) using local floral zoologist Centre and Brussels references (5, 6)

Primary food West Africa (7), Mali (8), West Africa (7) DR. Congo (10) and Ecuador (12), Tanzania (11), Kenya (15) Sri Lanka (16) and India Vietnam (18) and ASEAN composition Nigeria (9) Tanzania (11) Peru (13) and and West Africa (7) (17) (19) tables used Central America (14)

DR: Democratic Republic

* 74% was formally classified as urban (Kisangani city), however it was relying on local food supply due to post-war conflict and was considered essentially a rural food system during data collection

5 References Table S1

1. Akoègninou A, Van der Burg W, & van der Maese L (2006) Flore analytique du Bénin. (Wageningen, Netherlands), p 1034. 2. Bamps P (2000) Flore d'Afrique Centrale (Congo-Kinshasa, Rwanda, Burundi). (Meise). 3. de la Torre L, Navarrete H, Muriel P, Macia M, & Balslev H (2008) Enciclopedia de las plantas utiles del Ecuador. (Quito & Aarhus), pp 1-332. 4. Bridson D & Forman L (2016) The Herbarium Handbook. (London). 5. Assanayake MD (2003) A revised handbook to the flora of Ceylon. (Zeitltlinger publishers, New Zealand). 6. Ashton M, Gunatilleke S, Zoysa N, & M.D D (1997) A field guide to the common trees and shrubs of Sri Lanka. in Wild life Heritage Trust publications (Colombo). 7. Stadlmayr B, Charrondière R, Enujiugha V, Bayili R, Fagbohoun E , Samb B et al (2012) West African Food Composition Table. (Rome), p 148. 8. Barikmo I, Ouattara F, & Oshaug A (2004) Table de composition d'aliments du Mali. Research series No. 9. (Norway). 9. Oguntona E & Akinyele I (1995) Nutrient Composition of Commonly Eaten Foods in Nigeria - Raw, processed and prepared. in Publication Series 1995 (Ibadan, Nigeria), p 131. 10. De Groote V (1966) Tables de composition alimentaire pour la République Démocratique du Congo. (BP 3119 Kinshasa-Kaslina). 11. Lukmanji Z, et al. (2008) Tanzania Food Composition Tables. (Dar es Salaam and Boston). 12. Instituto Nacional de Nutricion (1965) Tabla de composicion de alimentos Ecuatorianos. Ministerio de provision social y sanidad (Quito). 13. Centro Nacional de Alimentacion y Nutricion (2009) Tablas Peruanas de composicion de alimentos. (Lima). http://www.ins.gob.pe/insvirtual/images/otrpubs/pdf/Tabla%20de%20Alimentos.pdf 14. INCAP (2007) Tabla de composicion de alimentos de centro america. Segunda Edicion. 15. Sehmi J (1993) National Food Composition Tables and The Planning of Satisfactory Diets in Kenya. (Nairobi, Kenya), p 200. 16. Thamilini J, Silva K, Sirasa M, & Samarasinghe W (2016) Food composition data in Sri Lanka: Past, present and future. in 11th international food data conference "Food composition and public health nutrition" (Hyderabad, India), pp S1-P33. 17. Gopalan C, Rama Sastri B, Balasubramanian, & S (1982) Nutritive Value of Indian Foods. (New Delhi, India), p 204. 18. National Institute of Nutrition (1972) Food products in Vietnam: composition and nutritive value. (Hanoi, Vietnam). 19. Institute of Nutrition Mahidol University (2014) ASEAN Food Composition Database, Electronic version 1. (Thailand). http://www.inmu.mahidol.ac.th/aseanfoods/composition_data.html

6 Table S2 Mean quantity consumed per food group (g/day)

Dry season Wet season Both seasons Food group (Benin, Kenya and (Benin, Kenya and (All countries)† Vietnam)* Vietnam)*

Mean± SD Mean± SD Mean± SD

All starchy staples 94.94±43.51 88.71±37.86 131.73±59.84

Pulses 136.14±19.21 118.68±32.01 92.34±40.53

Nuts and seeds 59.93±31.53 65.14±61.8 50.86±30.11

Dairy 81.14±45.03 75.39±44.8 63.71±52.53

Meat, poultry and fish‡ 40.99±39.89 41.83±39.04 44.68±24.01

Eggs 50.76±24.05 46.61±18.51 49.69±12.55

Dark green leafy vegetables 124.58±79.73 59.9±0.15 82.12±46.41

Vitamin A rich vegetables and fruits 50.78±44.51 33.09±25.68 47.91±24.24

Other vegetables 90.02±35.48 98.1±71.72 68.73±38.63

Other fruits 34.06±24.46 44.53±36.52 53.30±39.24

Not included in the DDS § 4.75±4.4 5.53±5.36 15.37±13.49

Food groups according to diet diversity score for women (42). The diet diversity score for children is based on 7 food groups (26) and aggregates “Pulses” and “Nuts and seeds” in one food group “Legumes and nuts”, “Dark green leafy vegetables” and “Other vitamin A-rich fruits and vegetables” into one food group”Vitamin A-rich fruits and vegetables” and “Other vegetables and “Other fruits” into one food group “Other fruits and vegetables”

DDS: Diet diversity score

* To enable comparison between dry and wet season only data for Benin, Kenya and Vietnam are tabulated

† Data from Benin, Cameroon, DR Congo, Ecuador, Kenya, Sri Lanka, Vietnam combined

‡ Flesh foods contain meat, poultry and fish but not insects or small protein food.

§ Foods not included in the DDS classification e.g. oils and fats, biscuits, insects, spices and condiments and foods consumed <15g by adult women

7 Table S3 Nutrient adequacy ratios of vitamin A, vitamin C, folate, calcium, iron and zinc in women and children by study area and season

Benin Cameroon Congo Ecuador Kenya Sri Lanka Vietnam All (n=2,439) (n=125) (n=462) (n=258) (n=1,580) (n=56) (n=1,306) (n=6,226)

Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD

Vitamin A

Dry season 0.28 0.29 NA NA NA NA NA NA 0.81 0.32 NA NA 0.88 0.28 0.62 0.31

Wet season 0.24 0.27 0.97 0.14 0.97 0.15 0.87 0.24 0.68 0.35 0.54 0.40 0.82 0.33 0.73 0.26

Vitamin C

Dry season 0.64 0.36 NA NA NA NA NA NA 0.88 0.27 NA NA 0.79 0.31 0.77 0.11

Wet season 0.62 0.38 0.83 0.28 0.95 0.16 0.97 0.11 0.83 0.31 0.65 0.34 0.82 0.30 0.81 0.13

Folate

Dry season 0.64 0.33 NA NA NA NA NA NA 0.80 0.27 NA NA 0.69 0.29 0.69 0.04

Wet season 0.63 0.33 0.72 0.27 0.64 0.29 NA NA 0.63 0.28 0.73 0.29 0.70 0.29 0.69 0.04

Calcium

Dry season 0.46 0.33 NA NA NA NA NA NA 0.57 0.31 NA NA 0.53 0.29 0.51 0.04

Wet season 0.40 0.32 0.36 0.26 0.52 0.32 0.49 0.26 0.47 0.32 0.48 0.31 0.60 0.31 0.48 0.08

Iron

Dry season 0.18 0.16 NA NA NA NA NA NA 0.16 0.10 NA NA 0.22 0.18 0.18 0.03

Wet season 0.17 0.16 0.16 0.13 0.10 0.07 0.15 0.15 0.15 0.11 0.09 0.18 0.28 0.28 0.17 0.06

Zinc

Dry season 0.58 0.32 NA NA NA NA NA NA 0.81 0.26 NA NA 0.97 0.11 0.78 0.19

Wet season 0.57 0.32 0.63 0.28 0.67 0.29 0.76 0.24 0.79 0.26 0.91 0.15 0.94 0.16 0.76 0.14 NA: Data not available

8 Table S4 Average number of species consumed per food group

Dry season Wet season Both seasons

(Benin, Kenya and (Benin, Kenya and (All countries) Vietnam) * Vietnam) * †

Mean± SD Mean± SD Mean± SD

All starchy staples 2.34±1.11 2.35±1.08 2.11±0.06

Pulses 1.12±0.33 1.03±0.18 1.14±0.03

Nuts and seeds 1.26±0.55 1.29±0.58 1.25±0.03

Dairy 1.00±0.00 1.01±0.08 1.00±0.00

Meat, poultry fish‡ 1.61±0.72 1.56±0.77 1.39±0.06

Eggs 1.02±0.14 1.06±0.23 1.01±0.00

Dark green leafy vegetables 1.46±0.74 1.54±0.8 1.35±0.17

Vitamin A-rich vegetables and fruits 2.14±0.82 1.98±0.76 1.7±0.32

Other vegetables 1.70±0.64 1.60±0.67 1.66±0.07

Other fruits 2.23±0.76 2.25±0.76 2.13±0.05

Not included in the DDS § 3.33±1.28 3.37±1.25 4.30±1.25

Food groups according to diet diversity score for women (1) . The Diet diversity score for children is based on 7 food groups (2) and aggregates “Pulses” and “Nuts and seeds” in one food group “Legumes and nuts”, “Dark green leafy vegetables” and “Other vitamin A-rich fruits and vegetables” into one food group”Vitamin A-rich fruits and vegetables” and “Other vegetables and “Other fruits” into one food group “Other fruits and vegetables”

DDS: Diet diversity score

* To enable comparison between dry and wet season only data for Benin, Kenya and Vietnam are tabulated

† Data from Benin, Cameroon, DR. Congo, Ecuador, Kenya, Sri Lanka, Vietnam combined

‡ Flesh foods contain meat, poultry and fish but not insects or small protein food.

§ Foods not included in the DDS classification e.g. oils and fats, biscuits, insects, spices and condiments or foods consumed <15g by women

References Table S4

1. FAO & FHI360 (2016) Minimum dietary diversity for women: A guide to measurement. (Rome). http://www.fao.org/3/a-i5486e.pdf 2. WHO (2008) Indicators for assessing infant and young child feeding practices. Part I : definitions. in Conclusions of a consensus meeting held 6-8 november 2007, Washington DC, USA (Geneva). http://apps.who.int/iris/bitstream/10665/43895/1/9789241596664_eng.pdf

9 Table S5 Species consumed by country and food group by women in 5 countries

DR. Congo Ecuador Kenya Sri Lanka Vietnam

All starchy Colocasia esculenta (L.) Schott Avena sativa L. Dioscorea spp. Colocasia esculenta (L.) Schott Alocasia macrorrhizos (L.) G. Don staples Dioscorea spp. Chenopodium quinoa Willd. Eleusine coracana (L.) Gaertn. Ipomoea batatas (L.) Lam. Centella asiatica (L.) Urb. Ipomoea batatas (L.) Lam. Hordeum vulgare L. Ipomoea batatas (L.) Lam. Manihot esculenta Crantz Colocasia esculenta (L.) Schott Manihot esculenta Crantz Manihot esculenta Crantz Manihot esculenta Crantz Oryza sativa L. Ipomoea batatas (L.) Lam. Musa sp. Oryza sativa L. Oryza sativa L. Solanum tuberosum L. Oryza sativa L. Oryza sativa L. Solanum tuberosum L. Solanum tuberosum L. Triticum aestivum L. Solanum tuberosum L. Solanum tuberosum L. Sorghum bicolor (L.) Moench Sorghum bicolor (L.) Moench Triticum turgidum L. Triticum aestivum L. Triticum aestivum L. Triticum aestivum L. Triticum aestivum L. Zea mays L. Zea mays L. Zea mays L. Ullucus tuberosus Caldas Zea mays L. Zea mays L. Pulses Glycine max (L.) Merr. Glycine max (L.) Merr. Glycine max (L.) Merr. Cicer arietinum L. Glycine max (L.) Merr. Phaseolus vulgaris L. Lens spp. Phaseolus vulgaris L. Curcuma longa L. Pachyrhizus erosus (L.) Urb. Vigna unguiculata (L.) Walp. Phaseolus lunatus L. Vigna radiata (L.) R. Wilczek Glycine max (L.) Merr. Phaseolus vulgaris L. Vicia faba L. Vigna unguiculata (L.) Walp. Lens culinaris Medik. Vigna radiata (L.) R. Wilczek Vigna unguiculata (L.) Walp. Vigna unguiculata (L.) Walp. Nuts and Arachis hypogaea L. Arachis hypogaea L. Arachis hypogaea L. NR Arachis hypogaea L. seeds Castanea sativa Mill. Sesamum indicum L. Dairy Bos taurus Linnaeus, 1758 NR Bos taurus Linnaeus, 1758 NR Bos taurus Linnaeus, 1758 Capra hircus Linnaeus, 1758 Meat, Bos taurus Linnaeus, 1758 Anas sp. Anas sp. Bos taurus Linnaeus, 1758 Anas platyrhynchos Linnaeus, 1758 poultry and Capra hircus Linnaeus, 1758 Bos taurus Linnaeus, 1758 Bos taurus Linnaeus, 1758 Gallus gallus Linnaeus, 1758 Anas sp. fish Cephalophus natalensis A. Smith, 1834 Cervidae spp. Gallus gallus Linnaeus, 1758 Scomberomorus lineolatus Cuvier, 1829 Anser anser Linnaeus, 1758 Gallus gallus Linnaeus, 1758 Gallus gallus Linnaeus, 1758 Rastrineobola argentea Sus scrofa Linnaeus, 1758 Bos Taurus Linnaeus, 1758 Brachyura spp. Rattus spp. Sardinops sagax Jenyns, 1842 Pellegrin, 1904 Thunnus albacares Bonnaterre, 1788 Bubalus bubalis Linnaeus, 1758 Sardina pilchardus Walbaum, 1792 Sus scrofa Linnaeus, 1758 Sus scrofa Linnaeus, 1758 Cairina moschata Linnaeus, 1758 Sardinops sagax Jenyns, 1842 Thunnus thynnus Linnaeus, 1758 Canis lupus Linnaeus, 1758 Sus scrofa Linnaeus, 1758 Ucides occidentalis Ortmann, 1897 Capra hircus Linnaeus, 1758 Centropus sinensis Stephens, 1815 Coturnix japonica Temminck & Schlegel, 1849 Ctenopharyngodon idella Valenciennes, 1844 Cyprinidae spp. Equus (Equus) caballus Linnaeus, 1758 Felis catus Linnaeus, 1778 Gallus gallus Linnaeus, 1758 Hoplobatrachus tigerinus Daudin, 1803 Meleagris spp. Perca fluviatilis Linnaeus, 1758 Piaractus brachypomus Cuvier, 1818 Sus scrofa Linnaeus, 1758 Sylvilagus spp. Tilapia spp.

10 DR. Congo Ecuador Kenya Sri Lanka Vietnam

Eggs Gallus gallus Linnaeus, 1758 Gallus gallus Linnaeus, 1758 Gallus gallus Linnaeus, 1758 NR Anas sp. Cairina moschata Linnaeus, 1758 Gallus gallus Linnaeus, 1758 Dark green Amaranthus cruentus L. Beta vulgaris L. Amaranthus cruentus L. Allium ampeloprasum L. Acacia pennata (L.) Willd. leafy Amaranthus dubius Mart. ex Thell. Lactuca sativa L. Beta vulgaris L. Centella asiatica (L.) Urb. Amaranthus cruentus L. vegetables Apium graveolens L. Spinacia oleracea L. Brassica oleracea L. Curcuma longa L. Basella alba L. Basella alba L. Cleome gynandra L. Brassica juncea (L.) Czern. Gnetum africanum Welw. Corchorus olitorius L. Carica papaya L. Hibiscus sabdariffa L. Coriandrum sativum L. Centella asiatica (L.) Urb. Ipomoea batatas (L.) Lam. Crotalaria spp. Citrus limon (L.) Osbeck Cucurbita maxima Duchesne Corchorus capsularis L. Ipomoea batatas (L.) Lam. Coriandrum sativum L. Phaseolus vulgaris L. Cucurbita maxima Duchesne Solanum americanum Mill. Eryngium foetidum L. Vigna unguiculata (L.) Walp. Glebionis coronaria (L.) Cass. ex Spach Houttuynia cordata Thunb. Ipomoea aquatica Forssk. Ipomoea batatas (L.) Lam. Lactuca sativa L. Manihot esculenta Crantz Marsilea quadrifolia L. Mentha spp. Nasturtium microphyllum (Boenn. ex Rchb.) Rchb. Oxypolis filiformis (Walter) Britton Piper sarmentosum Roxb. Sauropus androgynus (L.) Merr. Vitamin A Anonidium mannii (Oliv.) Engl. & Diels Beta vulgaris L. Carica papaya L. Beta vulgaris L. Carica papaya L. rich Carica papaya L. Carica papaya L. Cucurbita maxima Duchesne Capsicum annuum L. Cucurbita maxima Duchesne vegetables Elaeis guineensis Jacq. Carica x pentagona Heilborn Daucus carota L. Daucus carota L. Daucus carota L. and fruits Persea americana Mill. Daucus carota L. Ipomoea batatas (L.) Lam. Hylocereus undatus (Haw.) Britton & Rose Persea americana Mill. Persea americana Mill. Persea americana Mill. Sechium edule (Jacq.) Sw. Other Allium cepa L. Allium cepa L. Agaricus bisporus (J.E. Lange) Abelmoschus esculentus (L.) Moench Allium fistulosum L. vegetables Allium fistulosum L. Armoracia rusticana P. Gaertn., Imbach, 1946 Allium cepa L. Artocarpus altilis (Parkinson ex F.A. Zo) Brassica oleracea L. B.Mey. & Scherb. Allium cepa L. Artocarpus altilis (Parkinson ex F.A.Zo) Fosberg Megaphrynium macrostachyum (K. Brassica oleracea L. Brassica oleracea L. Fosberg Bambusa vulgaris Schrad. Schum.) Milne-Redh. Brassica rapa L. Crotalaria brevidens Benth. Brassica oleracea L. Benincasa hispida (Thunb.) Cogn. Saccharum officinarum L. Brassica spp. Saccharum officinarum L. Capsicum annuum L. Brassica oleracea L. Solanum aethiopicum L. Citrus limon (L.) Osbeck Solanum lycopersicum L. Lasia spinosa (L.) Thwaites Brassica rapa L. Solanum lycopersicum L. Citrus maxima (Burm.) Merr. Solanum lycopersicum L. Capsicum annuum L. Solanum melongena L. Cucumis sativus L. Solanum melongena L. Cucumis sativus L. Solanum lycopersicum L. Gigantochloa spp. Luffa cylindrica (L.) M. Roem. Physalis angulata L. Saccharum officinarum L.

11 DR. Congo Ecuador Kenya Sri Lanka Vietnam

Solanum anguivi Lam. Solanum lycopersicum L. Solanum macrocarpon L. Solanum melongena L. Vigna radiata (L.) R. Wilczek Other fruits Ananas comosus (L.) Merr. Ananas comosus (L.) Merr. Citrullus lanatus (Thunb.) Artocarpus altilis (Parkinson ex F.A.Zo) Averrhoa carambola L. Citrus limon (L.) Osbeck Annona muricata L. Matsum. & Nakai Fosberg Carica papaya L. Citrus reticulata Blanco Carica papaya L. Citrus x aurantium L. Cocos nucifera L. Citrullus lanatus (Thunb.) Matsum. & Nakai Citrus x aurantium L. Citrullus lanatus (Thunb.) Matsum. & Mangifera indica L. Luffa acutangula (L.) Roxb. Citrus limon (L.) Osbeck Cocos nucifera L. Nakai Musa sp. Malus pumila Mill. Citrus maxima (Burm.) Merr. Cucurbita maxima Duchesne Citrus reticulata Blanco Passiflora edulis Sims Momordica charantia L. Citrus reticulata Blanco Dacryodes edulis (G. Don) H.J. Lam Citrus sinensis (L.) Osbeck Psidium guajava L. Murraya koenigii (L.) Spreng. Citrus x aurantium L. Musa sp. Cucumis melo L. Pyrus spp. Spondias dulcis Parkinson Cucurbita pepo L. Syzygium aromaticum (L.) Merr. & L.M. Cyphomandra betacea (Cav.) Sendtn. Dimocarpus longan Lour. Perry Daucus carota L. Ficus racemosa L. Talinum fruticosum (L.) Juss. Fragaria spp. Lagenaria siceraria (Molina) Standl. Spondias dulcis Parkinson Gurania spp. Malus pumila Mill. Inga vera Willd. Musa sp. Malus pumila Mill. Psidium guajava L. Malus sylvestris (L.) Mill. Punica granatum L. Musa sp. Tamarindus indica L. Passiflora edulis Sims Ziziphus jujuba Mill. Passiflora quadrangularis L. Prunus domestica L. Prunus persica (L.) Batsch Psidium guajava L. Rubus caesius L. Solanum quitoense Lam. Syzygium aromaticum (L.) Merr. & L.M.Perry Tamarindus indica L. Vitis vinifera L. Not Aframomum laurentii (De Wild. & T. Allium cepa L. Agaricus bisporus (J.E. Lange) Allium cepa L. Allium fistulosum L. included in Durand) K. Schum. Allium sativum L. Imbach, 1946 Arachis hypogaea L. Allium sativum L. the DDS2 Allium cepa L. Anas sp. Allium cepa L. Bos taurus Linnaeus, 1758 Alpinia galanga (L.) Willd. Allium fistulosum L. Arachis hypogaea L. Allium sativum L. Capsicum annuum L. Amalocalyx microlobus Pierre ex Spire Allium porrum L. Armoracia rusticana P. Gaertn., Amaranthus cruentus L. Citrus limon (L.) Osbeck Anas sp. Allium sativum L. B.Mey. & Scherb. Arachis hypogaea L. Cocos nucifera L. Anethum graveolens L. Apium graveolens L. Avena sativa L. Bos taurus Linnaeus, 1758 Curcuma longa L. Arachis hypogaea L. Arachis hypogaea L. Bactris gasipaes Kunth Brassica oleracea L. Manihot esculenta Crantz Artocarpus altilis (Parkinson ex F.A.Zo) Bos taurus Linnaeus, 1758 Beta vulgaris L. Camellia sinensis (L.) Kuntze Murraya koenigii (L.) Spreng. Fosberg Capra hircus Linnaeus, 1758 Bixa orellana L. Citrus aurantiifolia (Christm.) Oryza sativa L. Averrhoa carambola L. Capsicum annuum L. Bos taurus Linnaeus, 1758 Swingle Osmundastrum cinnamomeum (L.) C. Bambusa vulgaris Schrad. Cinnamomum verum J. Presl Brassica oleracea L. Citrus limon (L.) Osbeck Presl Basella alba L. Coffea sp. Brassica rapa L. Coffea sp. Piper nigrum L. Bos taurus Linnaeus, 1758 Cucurbita maxima Duchesne Brassica spp. Corchorus olitorius L. Psophocarpus scandens (Endl.) Verdc. Brassica juncea (L.) Czern.

12 DR. Congo Ecuador Kenya Sri Lanka Vietnam

Dacryodes edulis (G. Don) H.J. Lam Capsicum annuum L. Coriandrum sativum L. Solanum lycopersicum L. Brassica oleracea L. Elaeis guineensis Jacq. Cervidae spp. Crotalaria brevidens Benth. Thunnus albacares Bonnaterre, 1788 Bubalus bubalis Linnaeus, 1758 Glycine max (L.) Merr. Citrus limon (L.) Osbeck. Cucurbita maxima Duchesne Triticum aestivum L. Cairina moschata Linnaeus, 1758 Helix pomatia Linnaeus, 1758 Citrus sinensis (L.) Osbeck Eleusine coracana (L.) Gaertn. Vigna radiata (L.) R.Wilczek Canis lupus Linnaeus, 1758 Hibiscus sabdariffa L. Coffea canephora Pierre ex Glycine max (L.) Merr. Zingiber officinale Roscoe Capra hircus Linnaeus, 1758 Hua gaboni Pierre ex De Wild. A.Froehner Ipomoea batatas (L.) Lam. Capsicum annuum L. Laurus nobilis L. Coriandrum sativum L. Manihot esculenta Crantz Carica papaya L. Manihot esculenta Crantz Cuminum cyminum L. Musa sp. Citrullus lanatus (Thunb.) Matsum. & Nakai Myristica fragrans Houtt. Cuniculus paca Linnaeus, 1766 Oryza sativa L. Citrus limon (L.) Osbeck Persea americana Mill. Daucus carota L. Phaseolus vulgaris L. Citrus maxima (Burm.) Merr. Piper guineense Schumach. & Thonn. Elaeis guineensis Jacq. Rastrineobola argentea Citrus x aurantium L. Rattus spp. Gallus gallus Linnaeus, 1758 Pellegrin, 1904 Colocasia esculenta (L.) Schott Saccharum officinarum L. Hordeum vulgare L. Saccharum officinarum L. Coriandrum sativum L. Sardina pilchardus Walbaum, 1792 Lactuca sativa L. Solanum americanum Mill. Crocothemis servilia Drury, 1773 Sardinops sagax Jenyns, 1842 Lens spp. Solanum lycopersicum L. Ctenopharyngodon idella Valenciennes, 1844 Scorodophloeus zenkeri Harms Manihot esculenta Crantz Solanum tuberosum L. Cucurbita maxima Duchesne Solanum lycopersicum L. Musa sp. Sorghum bicolor (L.) Moench Curcuma longa L. Solanum melongena L. Origanum vulgare L. Sus scrofa Linnaeus, 1758 Cymbopogon citratus (DC.) Stapf Triticum aestivum L. Oryza sativa L. Theobroma cacao L. Cyprinidae spp. Zea mays L. Osmundastrum cinnamomeum (L.) C. Triticum aestivum L. Dimocarpus longan Lour. Zingiber officinale Roscoe Presl Vigna subterranea (L.) Verdc. Enydra Fluctuans DC. Passiflora edulis Sims Vigna unguiculata (L.) Walp. Eryngium foetidum L. Phaseolus lunatus L. Zea mays L. Gallus gallus Linnaeus, 1758 Piper nigrum L. Zingiber officinale Roscoe Garcinia oblongifolia Champ. ex Benth. Pisum sativum L. Gigantochloa spp. Psidium guajava L. Glebionis coronaria (L.) Cass. ex Spach Saccharum officinarum L. Glycine max (L.) Merr. Solanum lycopersicum L. Gryllus spp. Solanum quitoense Lam. Helix spp. Solanum tuberosum L. Hoplobatrachus tigerinus Daudin, 1802 Sorghum bicolor (L.) Moench Ipomoea aquatica Forssk. Sus scrofa Linnaeus, 1758 Ipomoea batatas (L.) Lam. Syzygium aromaticum (L.) Merr. & Paederia foetida L. L.M. Perry Luffa cylindrica (L.) M. Roem. Theobroma cacao L. Malus pumila Mill. Thunnus thynnus Linnaeus, 1758 Marsilea quadrifolia L. Triticum aestivum L. Melissa officinalis L. Ullucus tuberosus Caldas Mentha spp. Vicia faba L. Ocimum basilicum L. Zea mays L. Oryza sativa L. Panax vietnamensis Ha & Grushv. Phaseolus vulgaris L. Physalis angulata L. Piper nigrum L. Piper sarmentosum Roxb. Polygonum odoratum (Mill.) Druce

13 DR. Congo Ecuador Kenya Sri Lanka Vietnam

Saccharum officinarum L. Sauropus androgynus (L.) Merr. Sechium edule (Jacq.) Sw. Sesamum indicum L. Solanum anguivi Lam. Solanum lycopersicum L. Solanum macrocarpon L. Solanum melongena L. Solanum tuberosum L. Sus scrofa Linnaeus, 1758 Tilapia spp. Triticum aestivum L. Vespa velutina Lepeletier, 1836 Vigna radiata (L.) R.Wilczek Zea mays L. Zingiber officinale Roscoe Table organised according to the food groups of the dietary diversity score for women (1). NR: Not reported; DDS: Diet diversity score * Flesh foods contain meat, poultry and fish but not insects or small protein food † Foods not included in the DDS classification e.g. oils, broth, biscuits, sweets, insects, spices or condiments and foods consumed <15g

Reference 1. FAO & FHI360 (2016) Minimum dietary diversity for women: A guide to measurement. (Rome). http://www.fao.org/3/a-i5486e.pdf

14 Table S6 Species consumed by country and food group by children in 5 countries

Benin Cameroon Kenya Sri Lanka Vietnam

All starchy staples Dioscorea spp. Colocasia esculenta (L.) Schott Dioscorea spp. Manihot esculenta Crantz Alocasia macrorrhizos (L.) G. Don Ipomoea batatas (L.) Lam. Dioscorea spp. Eleusine coracana (L.) Oryza sativa L. Centella asiatica (L.) Urb. Manihot esculenta Crantz Manihot esculenta Crantz Gaertn. Solanum tuberosum L. Colocasia esculenta (L.) Schott Musa sp. Musa sp. Ipomoea batatas (L.) Lam. Triticum aestivum L. Ipomoea batatas (L.) Lam. Oryza sativa L. Oryza sativa L. Manihot esculenta Crantz Triticum turgidum L. Oryza sativa L. Solanum tuberosum L. Solanum tuberosum L. Oryza sativa L. Zea mays L. Solanum tuberosum L. Sorghum bicolor (L.) Moench Triticum aestivum L. Solanum tuberosum L. Triticum aestivum L. Triticum aestivum L. Zea mays L. Sorghum bicolor (L.) Zea mays L. Triticum turgidum L. Moench Zea mays L. Triticum aestivum L. Zea mays L. Pulses, nuts and seeds Arachis hypogaea L. Arachis hypogaea L. Amaranthus cruentus L. Arachis hypogaea L. Arachis hypogaea L. Cajanus cajan (L.) Millsp. Citrullus lanatus (Thunb.) Matsum. & Arachis hypogaea L. Glycine max (L.) Merr. Glycine max (L.) Merr. Glycine max (L.) Merr. Nakai Glycine max (L.) Merr. Lens culinaris Medik. Pachyrhizus erosus (L.) Urb. Sesamum indicum L. Glycine max (L.) Merr. Phaseolus vulgaris L. Sesamum indicum L. Phaseolus vulgaris L. Vigna unguiculata (L.) Walp. Phaseolus vulgaris L. Vigna radiata (L.) R. Vigna unguiculata (L.) Walp. Sesamum indicum L. Ricinodendron heudelotii (Baill.) Hecke Wilczek Vigna radiata (L.) R. Wilczek Vigna unguiculata (L.) Walp. Vigna subterranea (L.) Vigna unguiculata (L.) Walp. Verdc. Vigna unguiculata (L.) Walp. Dairy Bos taurus Linnaeus, 1758 Bos taurus Linnaeus, 1758 Bos taurus Linnaeus, 1758 Bos taurus Linnaeus, 1758 Bos taurus Linnaeus, 1758 Capra hircus Linnaeus, Capra hircus Linnaeus, 1758 1758 Meat, poultry and fish Bos taurus Linnaeus, 1758 Bos taurus Linnaeus, 1758 Anas sp. Bos taurus Linnaeus, 1758 Anas platyrhynchos Linnaeus, 1758 Brachyura spp. Sus scrofa Linnaeus, 1758 Bos taurus Linnaeus, 1758 Gallus gallus Linnaeus, 1758 Anas sp. Chrysichthys nigrodigitatus Gallus gallus Linnaeus, Scomberomorus lineolatus Cuvier, 1829 Anser anser Linnaeus, 1758 Lacépède, 1803 1758 Sus scrofa Linnaeus, 1758 Bos taurus Linnaeus, 1758 Clarias spp. Rastrineobola argentea Thunnus albacares Bonnaterre, 1788 Bubalus bubalis Linnaeus, 1758 Decapterus punctatus Cuvier, Pellegrin, 1904 Cairina moschata Linnaeus, 1758 1829 Sus scrofa Linnaeus, 1758 Canis lupus Linnaeus, 1758 Ethmalosa fimbriata Bowdich, Capra hircus Linnaeus, 1758 1825 Coturnix japonica Temminck & Gallus gallus Linnaeus, 1758 Schlegel, 1849 Macrobrachyum spp. Ctenopharyngodon idella Valenciennes, Ovis aries Linnaeus, 1758 1844 Sardina pilchardus Walbaum, Cyprinidae spp. 1792 Equus (Equus) caballus Linnaeus, 1758 Sus scrofa Linnaeus, 1758 Felis catus Linnaeus, 1758 Sylvilagus spp. Gallus gallus Linnaeus, 1758 Tilapia spp. Hoplobatrachus tigerinus Daudin, 1802 Perca fluviatilis Linnaeus, 1758 Piaractus brachypomus Cuvier, 1818 Sus scrofa Linnaeus, 1758

15 Benin Cameroon Kenya Sri Lanka Vietnam

Sylvilagus spp. Tilapia spp. Eggs Gallus gallus Linnaeus, 1758 NR Gallus gallus Linnaeus, NR Anas sp. 1758 Cairina moschata Linnaeus, 1758 Gallus gallus Linnaeus, 1758 Vitamin A-rich fruits Amaranthus cruentus L. Amaranthus blitum L. Amaranthus cruentus L. Allium ampeloprasum L. Allium sativum L. and vegetables Bidens pilosa L. Baccharoides filigera (Oliv. & Hiern) Basella alba L. Beta vulgaris L. Amaranthus cruentus L. Capsicum annuum L. Isawumi, El-Ghazaly & B. Nord Beta vulgaris L. Capsicum annuum L. Anethum graveolens L. Carica papaya L. Cucurbita maxima Duchesne Brassica oleracea L. Centella asiatica (L.) Urb. Basella alba L. Cleome gynandra L. Daucus carota L. Carica papaya L. Daucus carota L. Brassica juncea (L.) Czern. Corchorus olitorius L. Gnetum africanum Welw. Cleome gynandra L. Manihot esculenta Crantz Carica papaya L. Daucus carota L. Ipomoea batatas (L.) Lam. Corchorus olitorius L. Psophocarpus scandens (Endl.) Verdc. Centella asiatica (L.) Urb. Manihot esculenta Crantz Ocimum gratissimum L. Coriandrum sativum L. Citrus limon (L.) Osbeck Moringa oleifera Lam. Solanum scabrum Mill. Crotalaria spp. Coriandrum sativum L. Solanum macrocarpon L. Vernonia amygdalina Delile Cucurbita maxima Cucurbita maxima Duchesne Solanum melongena L. Xanthosoma sagittifolium (L.) Schott Duchesne Daucus carota L. Telfairia occidentalis Hook.f. Daucus carota L. Diospyros kaki L.f. Vigna unguiculata (L.) Walp. Ipomoea batatas (L.) Lam. Eryngium foetidum L. Persea americana Mill. Glebionis coronaria (L.) Cass. ex Spach Phaseolus vulgaris L. Hylocereus undatus (Haw.) Britton & Solanum americanum Mill. Ros Vigna unguiculata (L.) Ipomoea aquatica Forssk. Walp. Ipomoea batatas (L.) Lam. Lactuca sativa L. Marsilea quadrifolia L. Melissa officinalis L. Mentha spp. Momordica cochinchinensis (Lour.) Spren Nasturtium microphyllum (Boenn. ex Rchb.) Rchb. Ocimum basilicum L. Oxypolis filiformis (Walter) Britton Persea americana Mill Piper sarmentosum Roxb. Sauropus androgynus (L.) Merr. Sechium edule (Jacq.) Sw. Zingiber officinale Roscoe

16 Benin Cameroon Kenya Sri Lanka Vietnam

Other fruits and Abelmoschus esculentus (L.) Abelmoschus esculentus (L.) Moench Agaricus bisporus (J.E. Allium cepa L. Allium fistulosum L. vegetables Moench Afrostyrax lepidophyllus Mildbr. Lange) Imbach, 1946 Artocarpus altilis (Parkinson ex F.A.Zo) Apium graveolens L. Allium cepa L. Allium cepa L. Allium cepa L. Fosberg Averrhoa carambola L. Ananas comosus (L.) Merr. Allium porrum L. Brassica oleracea L. Artocarpus heterophyllus Lam. Bambusa vulgaris Schrad. Artocarpus altilis (Parkinson ex Ananas comosus (L.) Merr. Citrullus lanatus (Thunb.) Brassica oleracea L. Benincasa hispida (Thunb.) Cogn. F.A.Zorn) Fosberg Brassica oleracea L. Matsum. & Nakai Citrus limon (L.) Osbeck Brassica oleracea L. Cajanus cajan (L.) Millsp. Musa sp. Citrus aurantiifolia Cocos nucifera L. Brassica rapa L. Capsicum annuum L. Passiflora edulis Sims (Christm.) Swingle. Lasia spinosa (L.) Thwaites Capsicum annuum L. Citrus limon (L.) Osbeck Psidium guajava L. Citrus x aurantium L. Luffa acutangula (L.) Roxb. Carica papaya L. Citrus x aurantium L. Solanum aethiopicum L. Crotalaria brevidens Benth. Malus pumila Mill. Citrullus lanatus (Thunb.) Matsum. & Cocos nucifera L. Solanum lycopersicum L. Mangifera indica L. Momordica charantia L. Nakai Musa sp. Solanum macrocarpon L. Musa sp. Murraya koenigii (L.) Spreng Citrus limon (L.) Osbeck Solanum lycopersicum L. Talinum fruticosum (L.) Juss. Passiflora edulis Sims Solanum lycopersicum L. Citrus maxima (Burm.) Merr. Solanum melongena L. Psidium guajava L. Solanum melongena L. Citrus reticulata Blanco Saccharum officinarum L. Citrus x aurantium L. Solanum lycopersicum L. Cucumis sativus L. Cucurbita pepo L. Dimocarpus longan Lour. Gigantochloa spp. Lagenaria siceraria (Molina) Standl. Luffa cylindrica (L.) M. Roem. Malus pumila Mill. Musa sp. Nephelium lappaceum L. Physalis angulata L. Prunus salicina Lindl. Psidium guajava L. Punica granatum L. Pyrus spp. Saccharum officinarum L. Solanum anguivi Lam. Solanum lycopersicum L. Solanum macrocarpon L. Solanum melongena L. Tamarindus indica L. Vigna radiata (L.) R. Wilczek Ziziphus jujuba Mill. Not included in the Allium sativum L. Allium sativum L. Allium sativum L. Capsicum annuum L. Allium sativum L. DDS † Arachis hypogaea L. Arachis hypogaea L. Bos taurus Linnaeus, 1758 Cocos nucifera L. Alpinia galanga (L.) Willd. Bos taurus Linnaeus, 1758 Capsicum annuum L. Brassica oleracea L. Curcuma longa L. Amalocalyx microlobus Pierre ex Spire Brassica nigra (L.) K. Koch Elaeis guineensis Jacq Camellia sinensis (L.) Osmundastrum cinnamomeum (L.) C. Arachis hypogaea L. Elaeis guineensis Jacq. Piper nigrum L. Kuntze Presl Bos taurus Linnaeus, 1758 Pimpinella anisum L. Saccharum officinarium L. Coffea sp. Piper nigrum L. Capsicum annuum L. Piper nigrum L. Solanum lycopersicum L. Corchorus olitorius L. Solanum lycopersicum L. Crocothemis servilia Drury, 1773 Saccharum officinarum L. Triticum aestivum L. Gallus gallus Linnaeus, Zingiber officinale Roscoe Cymbopogon citratus (DC.) Stapf Solanum lycopersicum L. Zingiber officinale Roscoe 1758 Enydra fluctuans DC.

17 Benin Cameroon Kenya Sri Lanka Vietnam

Zingiber officinale Roscoe Saccharum officinarum L. Glycine max (L.) Merr. Solanum americanum Mill. Gryllus spp. Theobroma cacao L. Helix spp. Triticum aestivum L. Piper nigrum L. Vigna unguiculata (L.) Piper sarmentosum Roxb. Walp. Sesamum indicum L. Zingiber officinale Roscoe Sus scrofa Linnaeus, 1758 Zingiber officinale Roscoe Table organised according to the food groups of the dietary diversity score for children (1). NR: Not reported; DDS: Diet diversity score * Flesh foods contain meat, poultry and fish but not insects or small protein food † Foods not included in the DDS classification e.g. oils, broth, biscuits, sweets, insects, spices or condiments

Reference

1. WHO (2008) Indicators for assessing infant and young child feeding practices. Part I : definitions. in Conclusions of a consensus meeting held 6-8 November 2007, Washington DC, USA (Geneva). http://apps.who.int/iris/bitstream/10665/43895/1/9789241596664_eng.pdf

18 Table S7: Area under the curve for various cut-offs of MAR MAR cut off DDS DSR x DDS

Adults

MAR 50% 0.79 0.79

MAR 60% 0.75 0.74

MAR 70% 0.69 0.68

MAR 80% 0.68 0.69

MAR 90% 0.65 0.64

Children

MAR 50% 0.66 0.64

MAR 60% 0.66 0.63

MAR 70% 0.66 0.63

MAR 80% 0.67 0.63

MAR 90% 0.68 0.60

MAR: Mean nutrient adequacy ratio; DDS: Diet diversity score, DSR: Diet species richness

19 Table S8 Test characteristics of cut-off values of DDS and DDSxDSR to define diets with MAR>50% in women

DDS DDS x DSR

Se Sp Cor Clas LR+ LR- Se Sp Cor Clas LR+ LR- ≥ 1 100.0% 0.0% 85.4% 1.00 100.0% 0.0% 85.4% 1.00 ≥ 2 98.1% 10.9% 85.4% 1.10 0.18 100.0% 1.0% 85.6% 1.01 0.00 ≥ 3 89.1% 39.6% 81.9% 1.48 0.28 100.0% 1.0% 85.5% 1.01 0.06 ≥ 4 72.6% 71.9% 72.5% 2.58 0.38 99.8% 2.2% 85.6% 1.02 0.10 ≥ 5 42.4% 95.9% 50.2% 10.20 0.60 99.5% 4.5% 85.7% 1.04 0.11 ≥ 6 15.7% 99.4% 27.9% 24.55 0.85 99.1% 6.7% 85.7% 1.06 0.13 ≥ 7 4.1% 100.0% 18.1% 0.96 98.2% 12.8% 85.8% 1.13 0.14 ≥ 8 0.7% 100.0% 15.1% 0.99 98.0% 14.4% 85.9% 1.15 0.14 ≥ 9 0.0% 100.0% 14.6% 1.00 97.1% 18.2% 85.6% 1.19 0.16 ≥ 10 96.6% 20.1% 85.5% 1.21 0.17 ≥ 12 95.3% 25.9% 85.2% 1.29 0.18 ≥ 14 93.5% 31.6% 84.5% 1.37 0.21 ≥ 15 92.4% 34.8% 84.0% 1.42 0.22 ≥ 16 91.0% 40.3% 83.6% 1.52 0.22 ≥ 18 89.9% 41.9% 82.9% 1.55 0.24 ≥ 20 86.0% 50.2% 80.8% 1.73 0.28 ≥ 21 84.3% 53.4% 79.8% 1.81 0.30 ≥ 22 81.8% 59.4% 78.6% 2.02 0.31 ≥ 24 81.4% 60.1% 78.3% 2.04 0.31 ≥ 25 75.7% 68.1% 74.6% 2.37 0.36 ≥ 26 75.5% 68.1% 74.5% 2.36 0.36 ≥ 27 75.4% 68.7% 74.4% 2.41 0.36 ≥ 28 73.4% 70.3% 72.9% 2.47 0.38 ≥ 30 69.1% 76.4% 70.2% 2.92 0.40 ≥ 32 67.0% 77.3% 68.5% 2.95 0.43 ≥ 33 60.8% 83.7% 64.2% 3.73 0.47 ≥ 34 60.1% 84.0% 63.6% 3.76 0.48 ≥ 35 60.0% 84.0% 63.5% 3.76 0.48 ≥ 36 58.6% 84.7% 62.4% 3.82 0.49 ≥ 39 52.2% 89.1% 57.6% 4.81 0.54 ≥ 40 52.1% 89.1% 57.5% 4.80 0.54 ≥ 42 45.5% 91.1% 52.1% 5.08 0.60 ≥ 44 44.7% 91.7% 51.6% 5.38 0.60 ≥ 45 42.4% 92.0% 49.6% 5.31 0.63 ≥ 48 36.6% 92.7% 44.7% 4.97 0.68

20 DDS DDS x DSR

Se Sp Cor Clas LR+ LR- Se Sp Cor Clas LR+ LR- ≥ 50 34.2% 94.9% 43.0% 6.68 0.69 ≥ 51 28.5% 95.5% 38.3% 6.38 0.75 ≥ 52 28.5% 95.9% 38.3% 6.86 0.75 ≥ 54 27.5% 96.5% 37.6% 7.83 0.75 ≥ 55 26.6% 96.5% 36.8% 7.56 0.76 ≥ 56 24.5% 96.8% 35.0% 7.67 0.78 ≥ 60 23.6% 97.8% 34.4% 10.55 0.78 ≥ 63 19.7% 98.4% 31.2% 12.34 0.82 ≥ 64 19.6% 98.4% 31.1% 12.27 0.82 ≥ 65 19.0% 99.0% 30.7% 19.83 0.82 ≥ 66 18.1% 99.0% 29.9% 18.92 0.83 ≥ 68 16.4% 99.0% 28.4% 17.10 0.84 ≥ 70 16.2% 99.0% 28.3% 16.88 0.85 ≥ 72 15.0% 99.4% 27.3% 23.44 0.86 ≥ 75 13.5% 99.4% 26.0% 21.14 0.87 ≥ 76 13.0% 99.4% 25.6% 20.37 0.88 ≥ 77 12.9% 99.4% 25.5% 20.20 0.88 ≥ 78 12.8% 99.4% 25.4% 19.95 0.88 ≥ 80 12.2% 99.7% 24.9% 38.02 0.88 ≥ 84 10.7% 99.7% 23.6% 33.41 0.90 ≥ 85 10.2% 99.7% 23.2% 31.88 0.90 ≥ 90 9.5% 99.7% 22.7% 29.83 0.91 ≥ 91 8.3% 99.7% 21.6% 26.08 0.92 ≥ 95 8.2% 99.7% 21.5% 25.57 0.92 ≥ 96 7.7% 99.7% 21.1% 24.04 0.93 ≥ 98 6.8% 99.7% 20.3% 21.31 0.93 ≥ 100 6.5% 99.7% 20.1% 20.46 0.94 ≥ 102 6.3% 99.7% 19.9% 19.61 0.94 ≥ 104 5.5% 100.0% 19.2% 0.95 ≥ 105 5.3% 100.0% 19.1% 0.95 ≥ 108 5.0% 100.0% 18.8% 0.95 ≥ 110 4.1% 100.0% 18.1% 0.96 ≥ 112 4.1% 100.0% 18.1% 0.96 ≥ 114 3.4% 100.0% 17.5% 0.97 ≥ 119 3.1% 100.0% 17.2% 0.97 ≥ 120 2.5% 100.0% 16.7% 0.98 ≥ 126 2.0% 100.0% 16.2% 0.98

21 DDS DDS x DSR

Se Sp Cor Clas LR+ LR- Se Sp Cor Clas LR+ LR- ≥ 128 1.6% 100.0% 15.9% 0.98 ≥ 132 1.5% 100.0% 15.9% 0.98 ≥ 133 1.4% 100.0% 15.8% 0.99 ≥ 136 1.2% 100.0% 15.6% 0.99 ≥ 138 1.1% 100.0% 15.5% 0.99 ≥ 140 1.0% 100.0% 15.5% 0.99 ≥ 147 0.7% 100.0% 15.2% 0.99 ≥ 154 0.5% 100.0% 15.0% 1.00 ≥ 160 0.4% 100.0% 14.9% 1.00 ≥ 161 0.4% 100.0% 14.9% 1.00 ≥ 168 0.3% 100.0% 14.8% 1.00 ≥ 176 0.2% 100.0% 14.8% 1.00 ≥ 184 0.1% 100.0% 14.7% 1.00 > 184 0.0% 100.0% 14.6% 1.00

DDS: Diet diversity score, DSR: Diet species richness, MAR: Mean nutrient adequacy, Se: Sensitivity, Sp: Specificity, Cor clas: % correct classification, LR+: Likelihood ratio for positive test results, LR-: Likelihood ratio for negative test results

22 Table S9 Test characteristics of cut-off values of DDS and DSRxDDS to define diets with MAR>50% in children

DDS DDSx DSR Se Sp Cor Clas LR+ LR- Se Sp Cor Clas LR+ LR- ≥ 1 100.0% 0.0% 56.9% 1.00 100.0% 0.0% 56.9% 1 ≥ 2 99.7% 6.3% 59.4% 1.06 0.05 99.9% 3.7% 58.4% 1.04 0.02 ≥ 3 97.1% 12.4% 60.6% 1.11 0.24 99.9% 5.4% 59.1% 1.06 0.02 ≥ 4 83.7% 35.2% 62.8% 1.29 0.46 99.9% 5.9% 59.3% 1.06 0.02 ≥ 5 42.2% 79.8% 58.4% 2.09 0.72 99.4% 7.3% 59.7% 1.07 0.08 ≥ 6 5.9% 99.4% 46.2% 9.30 0.95 99.4% 7.4% 59.7% 1.07 0.08 ≥ 7 0.6% 100.0% 43.5% 0.99 98.9% 8.9% 60.0% 1.08 0.13 ≥ 8 0.0% 100.0% 43.1% 1.00 98.9% 8.9% 60.1% 1.09 0.13 ≥ 9 98.5% 9.8% 60.3% 1.09 0.15 ≥ 10 97.2% 11.4% 60.2% 1.10 0.25 ≥ 12 96.5% 12.7% 60.4% 1.11 0.27 ≥ 14 93.1% 18.1% 60.7% 1.14 0.38 ≥ 15 93.1% 18.4% 60.9% 1.14 0.38 ≥ 16 91.2% 21.7% 61.2% 1.17 0.40 ≥ 18 89.5% 22.8% 60.7% 1.16 0.46 ≥ 20 87.6% 26.0% 61.0% 1.18 0.48 ≥ 21 84.3% 28.8% 60.4% 1.18 0.55 ≥ 24 82.6% 32.5% 61.0% 1.22 0.54 ≥ 25 77.0% 39.7% 60.9% 1.28 0.58 ≥ 27 76.2% 39.8% 60.5% 1.27 0.60 ≥ 28 75.3% 41.4% 60.7% 1.29 0.60 ≥ 30 71.3% 46.0% 60.4% 1.32 0.62 ≥ 32 69.0% 47.6% 59.8% 1.32 0.65 ≥ 33 64.4% 53.2% 59.6% 1.38 0.67 ≥ 35 64.3% 53.3% 59.6% 1.38 0.67 ≥ 36 62.0% 54.3% 58.7% 1.36 0.70 ≥ 39 55.9% 63.2% 59.1% 1.52 0.70 ≥ 40 55.9% 63.2% 59.0% 1.52 0.70 ≥ 42 46.5% 73.7% 58.2% 1.77 0.73 ≥ 44 46.0% 73.7% 58.0% 1.75 0.73 ≥ 45 40.8% 79.4% 57.5% 1.98 0.75 ≥ 48 37.6% 81.3% 56.4% 2.00 0.77 ≥ 49 33.5% 83.9% 55.2% 2.08 0.79 ≥ 50 33.4% 83.9% 55.2% 2.08 0.79 ≥ 52 28.3% 87.8% 54.0% 2.31 0.82

23 DDS DDSx DSR Se Sp Cor Clas LR+ LR- Se Sp Cor Clas LR+ LR- ≥ 54 26.9% 88.2% 53.3% 2.27 0.83 ≥ 55 26.3% 88.2% 53.0% 2.22 0.84 ≥ 56 20.6% 92.1% 51.5% 2.61 0.86 ≥ 60 19.8% 92.2% 51.1% 2.56 0.87 ≥ 63 13.1% 96.0% 48.9% 3.31 0.90 ≥ 64 13.0% 96.0% 48.9% 3.29 0.91 ≥ 65 13.0% 96.0% 48.8% 3.28 0.91 ≥ 66 8.6% 98.1% 47.2% 4.53 0.93 ≥ 68 8.2% 98.2% 47.0% 4.43 0.94 ≥ 70 8.1% 98.2% 47.0% 4.41 0.94 ≥ 72 5.4% 99.0% 45.8% 5.23 0.96 ≥ 75 5.0% 99.2% 45.6% 6.18 0.96 ≥ 77 3.3% 99.7% 44.9% 9.49 0.97 ≥ 78 3.2% 99.7% 44.8% 9.36 0.97 ≥ 80 2.5% 99.7% 44.4% 8.65 0.98 ≥ 84 2.2% 99.8% 44.3% 9.68 0.98 ≥ 85 1.7% 99.9% 44.1% 29.59 0.98 ≥ 90 1.4% 100.0% 44.0% 0.99 ≥ 95 1.0% 100.0% 43.7% 0.99 ≥ 96 1.0% 100.0% 43.7% 0.99 ≥ 98 0.5% 100.0% 43.4% 0.99 ≥ 102 0.4% 100.0% 43.3% 1.00 ≥ 108 0.2% 100.0% 43.2% 1.00 ≥ 112 0.1% 100.0% 43.2% 1.00 > 112 0.0% 100.0% 43.1% 1.00

DDS: Diet diversity score, DSR: Diet species richness, MAR: Mean nutrient adequacy, Se: Sensitivity, Sp: Specificity, Cor clas: % correct classification, LR+: Likelihood ratio for positive test results, LR-: Likelihood ratio for negative test results

24