Ex-ante economic impact assessment of introducing cassava or maize biofortified with pro-vitamin A in four Latin American countries M.Sc. thesis Anne Balling Jacobsen (HEK 05032) Department of Human Nutrition Faculty of Life Sciences, University of Copenhagen, Denmark 2008 Title of M.Sc. thesis Ex-ante economic impact assessment of introducing cassava or maize biofortified with pro-vitamin A in four Latin American countries Author Anne Balling Jacobsen Viktoriagade 8 D, 2. th 1655 Copenhagen V email: anneballing81@yahoo.dk Supervisors Dr. Shakuntala Haraksingh Thilsted and Dr. Henrik Friis Department of Human Nutrition, Faculty of Life Sciences University of Copenhagen, Denmark Dr. Per Pinstrup-Andersen, H. E. Babcock Professor Division of Nutritional Sciences, Cornell University, USA Front page image sources: Far left, upper corner, CIAT logo. Downloaded September 2008 from: webpc.ciat.cgiar.org:8091/FDRH/images/logo.jpg Far left, middle position, Latin American map. Downloaded and modified September 2008 from: http://vivirlatino.com/i/2008/01/LANDING-MAP-LATIN-AMERICA.jpg Far left, lower corner, AgroSalud logo. Downloaded September 2008 from: http://www.agrosalud.org/index.php?option=com_docman&task=cat_view&gid=12&Itemid=30 Far right middle position, eye infected with Bitot’s spots. Downloaded and modified September 2008 from: http://www.cehjournal.org/images/ts040006.jpg Far right lower position, cassava. Downloaded September 2008 from: http://edge.rit.edu/content/P07403/public/cassava.jpg Mid lower position, maize. Downloaded September 2008 from: http://www.thebaitwarehouse.co.uk/images/FRENCH%20MAIZE%201.jpg Table of contents LIST OF T ABLES AND FIGURES I PREFACE II SUMMARY III RESUMÉ IV ABBREVIATIONS V 1 INTRODUCTION 1 1.1 BACKGROUND 1 1.2 OBJECTIVES 2 1.3 PROJECT FRAMEWORK 3 1.3.1 LITERATURE SEARCH 3 2 THE MAGNITUDE OF VITAMIN A DEFICIENCY (VAD) 4 2.1 THE BURDEN OF VAD 4 2.1.1 GLOBAL STATUS 4 2.1.2 STATUS IN LATIN AMERICA AND THE CARIBBEAN 5 2.1.3 STATUS IN COLOMBIA, HONDURAS, MEXICO AND NICARAGUA 5 2.2 MAJOR VAD DISABILITIES AND OUTCOMES 6 2.2.1 XEROPHTHALMIA (X) 6 2.2.2 MORTALITY 8 3 CAUSES OF VAD 9 3.1 IMMEDIATE CAUSES 9 3.2 UNDERLYING CAUSES 9 3.3 ASSESSMENT OF VAD 11 4 STRATEGIES TO ELIMINATE VAD 14 4.1 VIT AMIN A (VA) SUPPLEMENT ATION 14 4.2 FORTIFICATION OF COMMERCIAL FOODS WITH VA 15 4.3 DIETARY DIVERSIFICATION 16 4.3.1 NATURAL FOOD SOURCES OF VA 16 4.3.2 CAROTENOID BIOCONVERSION 17 4.3.3 FACTORS AFFECTING THE BIOEFFICACY OF PRO-VA CAROTENOIDS 18 5 BIOFORTIFICATION 20 5.1 RATIONALE FOR CROSS-BREEDING CASSAVA AND MAIZE FOR A HIGHER PRO-VA CONTENT 20 5.2 THE BENEFICIARY SYSTEM OF RESEARCH & DEVELOPMENT (R&D): IMPROVING VA ST ATUS AND HEAL TH OUTCOMES 20 5.3 BIOEFFICACY OF BIOFORTIFIED ST APLE CULTIVARS 22 5.3.1 GERBIL STUDIES OF BIOFORTIFIED COMPONENTS DISSOLVED IN OIL 22 5.3.2 HUMAN CONSUMPTION STUDIES OF PRO-VA-BIOFORTIFIED CULTIVARS 23 I 6 METHODOLOGIES 24 6.1 CONCEPTUAL FRAMEWORK OF THE DISABILITY ADJUSTED LIFE YEARS (DALYS) FORMULA 24 6.2 SELECTION CRITERIA OF TARGET COUNTRIES 25 6.2.1 POPULATION AND DEMOGRAPHIC DATA 26 6.2.2 HEALTH DATA 26 6.2.3 NUTRITIONAL DATA 28 6.3 TARGET CULTIVARS ASSUMPTIONS 28 6.3.1 OPTIMISTIC AND PESSIMISTIC SCENARIOS 29 6.3.2 FINANCIAL INVESTMENTS 29 6.4 SENSITIVITY ANALYSIS PERFORMANCE 29 7 EX-ANTE IMPACT RESUL TS FOR COLOMBIA, HONDURAS, MEXICO AND NICARAGUA 30 7.1 CURRENT BURDEN OF VAD 30 7.2 POTENTIAL HEAL TH AND ECONOMIC IMPACT OF BIOFORTICATION WITH PRO-VA 30 7.2.1 DALYS GAINED DUE TO BIOFORTIFIED CASSAVA 30 7.2.2 DALYS GAINED DUE TO BIOFORTIFIED MAIZE 31 7.2.3 NEW INCIDENCE OF FUNCTIONAL OUTCOMES OF VAD 32 7.3 POTENTIAL COST-EFFECTIVENESS OF BIOFORTIFIED CASSAVA AND MAIZE 34 7.4 SENSITIVITY TO SELECTED PARAMETERS OF THE DALYS MODEL 35 7.4.1 CHILDHOOD X ESTIMATIONS 35 7.4.2 BIOCONVERSION OF BETA-CAROTENE AND BETA-CRYPTOXANTHIN 36 7.4.3 INCOME LEVELS OF TARGET POPULATION 36 7.4.4 DIETARY VA INTAKE IN COLOMBIA 37 8 DISCUSSION 38 8.1 VALIDITY AND INFLUENCE OF DATA APPLIED IN THED ALYS CALCULATIONS 38 8.1.1 ESTIMATIONS OF CHILDHOOD X 38 8.1.2 DIETARY INTAKE OF VA 39 8.1.3 BIOCONVERSION RATES FOR THE CAROTENOIDS IN CASSAVA AND MAIZE 40 8.1.4 INCOME LEVELS OF THE TARGET POPULATIONS 41 8.2 UNDERSTANDING AND INTERPRETING THE RESULTS FROM THED ALYS MODEL 42 8.2.1 OUTCOMES OF CURRENT ANALYSES IN RELATION TO OTHERD ALYS-BASED ASSESSMENTS 43 8.2.2 COMPARATIVE COSTS OF BIOFORTIFICATION IN RELATION TO CURRENTV AD STRATEGIES 43 8.2.3 NATIONAL OR REGIONAL APPROACH: EFFECT OF AGGREGATE INTAKE DATA 44 8.2.4 THE BENEFICIAL SCOPE OF THE DALYS RESULTS 45 8.3 FACTORS THAT MAY INFLUENCE THE ACHIEVEMENT OF THE CALCULATED IMPACT S 46 8.3.1 POLITICAL COMMITMENT TO ENSURE ADOPTION 47 8.3.2 CULTURE, FOOD HABITS AND EFFICACY 48 8.3.3 CO-PRESENCE OF OTHER NUTRITIONAL DEFICIENCIES 49 9 CONCLUSION 51 10 FUTURE PERSPECTIVES 52 11 ACKNOWLEDGEMENTS 53 12 REFERENCES 55 II APPENDIX A: APPLIED PREVALENCE OF VAD OUTCOMES 68 APPENDIX B: APPLIED DATA ON VA INT AKE 69 APPENDIX C: DATA SOURCES FOR THE DALYS CALCULATIONS 70 APPENDIX D: ASSUMPTIONS RELATED TO THE TWO SCENARIOS 74 APPENDIX E: EFFICACY T ABLES 76 APPENDIX F: NEW INCIDENCE RATES 82 APPENDIX G: SENSITIVITY RESUL TS OF THE EXTRAPOLATED <5 XN 84 APPENDIX H: SENSITIVITY OF BIOCONVERSION 85 APPENDIX I: SENSITIVITY TO INCOME LEVELS 88 APPENDIX J: SENSITIVITY OF DIET ARY VA INT AKE OF WOMEN 19-49 YEARS IN COLOMBIA 90 III List of tables and figures Tables Table 1. Country prevalence of V AD and estimated absolute number affected in Colombia, Honduras, Mexico and Nicaragua Table 2. X classification according to severity Table 3. Indicators of developmental and health character in Colombia, Honduras, Mexico and Nicaragua Table 4. Current VAD strategies and their coverage in Colombia, Honduras, Mexico and Nicaragua Table 5. Major factors affecting bioavailability of dietary carotenoids Table 6. DALYs lost to V AD per year in Colombia, Honduras, Mexico and Nicaragua Table 7. Gains in DALYs, USD and reduction of the health burden with biofortified cassava per year Table 8. Gains in DALYs, USD and reduction of the health burden with biofortified maize per year Table 9. Averted mortality in children <5 years of age per year due to biofortified cassava or maize Table 10. Number of deaths averted in children <5 years of age per 1,000,000 due to biofortified cassava or maize Table 11. Cost and gain in USD per DALY saved by biofortified cassava and maize Table 12. IRR and B/C over a 20 year period for biofortified cassava and maize Figures Figure 1. Prevalence of V AD in children <5 years of age Figure 2. Immediate factors contributing to the development of V AD, and the relation between major functional outcomes of this condition Figure 3. Underlying factors contributing to the development of undernutrition and infection Figure 4. Overview of Latin America and the Caribbean with the four countries under investigation Figure 5. Current dietary vitamin A consumption of women of reproductive age and preschool children in Colombia, Honduras, Mexico and Nicaragua Figure 6. Chemical structure of all-trans-retinol, all-trans-β-carotene, all-trans-α-carotene and all-trans-β- cryptoxanthin Figure 7. Technical beneficiary cycle of biofortification, and factors affecting the potential health and monetary impacts Figure 8. Flow diagram on the availability of collected secondary data critical to the DALYs model I Preface This M.Sc. thesis represents a strong collaborative effort by my supervisors and Dr. Helena Pachón, AgroSalud, Centro Internacional de Agricultura Tropical (CIAT). It constitutes a small part of the work launched by AgroSalud to determine which Latin American and Caribbean countries should be targeted for introduction of bean, cassava, maize, rice, or sweet potato biofortified with either zinc, iron, protein or vitamin A (AgroSalud 2008). Danish organizations awarded me funding to carry out work for my thesis, which began with a four months’ stay at CIAT, Cali, Colombia. During that stay, technical expertise to conduct the ex-ante calculations was provided on site, and AgroSalud financed a trip for me to go to Costa Rica, where I got the opportunity to present my work at the 54 th annual meeting of “el Programa Cooperativo Centroamericano para el Mejoramiento de Cultivos y Animales” (PCCMCA), May 2008. Work for the last part of my thesis was carried out in Denmark, and was strictly theoretical. During my stay in Colombia, I did not engage in fieldwork in relation to this thesis. _____________________________ Anne Balling Jacobsen, Copenhagen II Summary Background: In Latin America, vitamin A Deficiency (V AD) is considered subclinical despite preventive efforts of vitamin A supplementation and fortification. It is unknown whether biofortification of staple cultivars with pro-vitamin A can be a sustainable, complementary effort in eradicating V AD in countries in this region. Objectives: To assess the current health and economic burden of V AD and the potential impact of introducing pro-V A-biofortified cassava or maize in Colombia, Honduras, Mexico and Nicaragua. Methods: The analytical framework of Disability Adjusted Life Years (DALYs) was applied, in which per capita gross national product reflected the value of each year lost or saved from morbidity and mortality. DALYs calculations were carried out in two scenarios, applying optimistic and pessimistic assumptions about factors with respect to efficacy and coverage rate. Results: Current health costs of V AD appeared to range from USD 4.7 million in Nicaragua to USD 509.8 million in Mexico. The DALYs saved with the optimistic scenario ranged from 140 in Mexico to 1,815 in Nicaragua with biofortified cassava, and from 259 in Colombia to 10,291 in Mexico for biofortified maize. These improvements were mainly achieved through averting deaths of children <5 years of age. Internal Rates of Return were consistently higher for introducing biofortified maize compared to cassava in all four countries, ranging from 17% in Colombia to 216% in Mexico in the pessimistic and optimistic scenario, respectively. Conclusions: From a cost-benefit perspective, introduction of these pro-V A-biofortified cultivars stands to benefit all four countries in the long-term with the exception of the pessimistic scenario with cassava in Colombia. The levels of improvements did not appear adequate to solve the problems related to V AD in these countries. More investigations are needed to elucidate issues related to efficacy and adoption of biofortified cassava and maize in order to carry out more precise ex-ante assessments. III Resumé Baggrund: A-vitamin mangel betragtes i Latinamerika som subklinisk trods forebyggende tiltag, herunder vitamin A (V A) supplementering og fortificering. Det er uvist, hvorvidt biofortificering af basisafgrøder med pro-V A kan være en bæredygtig, komplementær løsning i udryddelsen af A-vitamin mangel i lande fra denne region. Formål: At estimere de nuværende sundhedsmæssige og økonomiske konsekvenser af A-vitamin mangel, og de potentielle indvirkninger af at introducere pro-V A- biofortificeret kassava eller majs, i Colombia, Honduras, Mexico og Nicaragua. Metoder: En analytisk model bygget på konceptet Disability Adjusted Life Years (DALYs) blev anvendt med brutto national produkt per indbygger svarende til værdien af hvert år tabt eller vundet på grund af morbiditet og mortalitet. DALYs kalkulationer blev udført i to scenarier som benyttede optimistiske og pessimistiske antagelser omkring faktorer relaterede til virkningskraften og dækningsgraden af de biofortificerede afgrøder. Resultater: Nuværende sundhedsomkostninger foranlediget af V AD synes at rangere fra USD 4,7 millioner i Nicaragua til USD 509,8 millioner i Mexico. DALYs vundet med det optimistiske scenario rangerede fra 140 i Mexico til 1.815 i Nicaragua med biofortificeret kassava, og fra 259 i Colombia til 10.291 i Mexico for biofortificeret majs. Disse forbedringer blev hovedsageligt opnået ved at afværge dødsfald iblandt børn <5 år. Internal Rates of Return var konsekvent højere ved at introducere biofortificeret majs sammenlignet med cassava i alle fire lande, hvilken rangerede fra 17% i Colombia til 216% i Mexico i henholdsvis det pessimistiske og optimistiske scenario. Konklusioner: Fra et cost-benefit-perspektiv ser introduktionen af de biofortificerede afgrøder ud til at gavne alle fire lande i det langsigtede perspektiv, med undtagelse af det pessimistiske scenario for kassava i Colombia. Forbedringsniveauerne syntes generelt ikke tilstrækkelige til at løse de problemer der er relaterede til A-vitamin mangel i de fire lande. Der er behov for flere undersøgelser for at udrede usikkerheder omkring virkningskraft og adoption af biofortificeret cassava og majs, så mere præcise ex-ante vurderinger kan udføres. IV Abbreviations B/C Benefit/Cost CBR Crude birth rate CGIAR Consultative Group on International Agricultural Research CIAT Centro Internacional de Agricultura Tropical (International Center for Tropical Agriculture) CIMMYT Centro Internacional de Mejoramiento de Maíz y Trigo (International Maize and Wheat Improvement Center) DALY Disability adjusted life year DHS Demographic and Health Survey DRD Deuterated retinol isotope dilution EAR Estimated average requirement FAO Food and Agriculture Organization FBS Food Balance Sheets GNP Gross National Product IFPRI International Food Policy Research Institute IOM Institute of Medicine IRR Internal Rates of Return IU International Unit IV ACG International Vitamin A Consultancy Group LSMS Living Standard Measurement Survey MDG Millennium Development Goal MRDR Modified relative dose response test NGO Non-governmental organization OFSP Orange-fleshed sweet potato PAHO Pan American Health Organization PEM Protein-energy malnutrition PRAF The Family Allowance Program PT Pupillary threshold R&D Research & Development RAE Retinol activity equivalents RDA Recommended dietary allowance RDR Relative dose response test SLE Standard life expectancy SES Socioeconomic status U5MR Under 5 mortality rate UNICEF United Nations Children’s Fund V A Vitamin A V AD Vitamin A deficiency V AS Vitamin A supplementation VMNIS Vitamin and Mineral Nutrition Information System WHO World Health Organization X Xerophthalmia X1B Bitot’s spots X3 Keratomalacia XS Corneal scars XN Night blindness V 1 Introduction 1.1 Background Eliminating Vitamin A deficiency (V AD) is an essential part of reducing childhood mortality, improving maternal health and, due to its association with blindness, assuring universal primary education, thus, aiding in three of the eight Millennium Development Goals (MDGs) (SCN 2004). Recent estimates based on the Disability Adjusted Life Years (DALYs) approach show that deficiency of Vitamin A (V A) is responsible for 5.3% of the global childhood years lost to disability, and it accounts for 0.6 million deaths per year (Black et al. 2008). Hence, V AD constitutes a major health problem globally warranting calls for action. Despite improvements in nutritional status attained partly by public intervention programs of supplementation and fortification with V A during the last decades, V AD is still present in some parts of Latin America and the Caribbean, although mostly considered subclinical (Mason et al. 2005; Mora, Gueri, & Mora 1998). Each year, an estimated 8.2 million preschool children and 375,000 women of reproductive age suffer from V AD, representing 17.3% and 3.8% of these population groups, respectively. However, when comparing the estimated prevalence of clinical V AD such as the ocular condition, xerophthalmia (X), maternal V AD seems to comprise the largest burden, affecting 376,000 women annually versus 20,000 preschool children (West 2002). Thus, the current burden of V AD in some Latin American countries still holds the potential for further improvement. Ongoing Research & Development (R&D) efforts for improving the nutritional content of traditional staple foods with cross-breeding techniques, a process known as biofortification (Nestel et al. 2006), may have the potential to improve nutritional status, and hence, health and productivity of disadvantaged people dependent on agriculture in developing countries. It is argued that introducing commonly consumed staples biofortified with pro-V A can serve as a sustainable, complementary approach in the efforts of eradicating V AD. However, for investments and implementation of national 1 policies, it is crucial to conduct ex-ante assessments of potential impact and cost- effectiveness of this food-based strategy and present these to national governments as well as non-governmental organizations (NGOs). For this purpose, the analytical framework based on the DALYs formula, which allows for quantification of current and future potential health burden per year on the basis of functional outcomes of V AD, may prove to be a practical tool. In Colombia, Honduras, Mexico, and Nicaragua, current efforts to prevent V AD include vitamin A supplementation (V AS) and fortification of commercial staple foods with V A (Darnton-Hill & Nalubola 2002). However, while V AD is virtually eradicated in Nicaragua, Mexico fights with extensive subclinical V AD on a scale that makes it a public health problem, suggesting that while effective in some countries, current interventions may only hold part of the solution for other countries. 1.2 Objectives The objectives of this M.Sc. thesis are as follows: To assess the national health and economic burden of V AD in Colombia, Honduras, Mexico and Nicaragua To assess the health and economic impact of introducing pro-V A-biofortified cassava or maize in Colombia, Honduras, Mexico and Nicaragua, and to evaluate the cost-effectiveness of this food-based strategy To evaluate the sensitivity of selected parameters in the DALYs calculations To evaluate the potential of introducing of pro-V A-biofortified cassava or maize as an alternative food-based strategy for alleviating V AD in the four countries under investigation 2 1.3 Project framework Since 2004, R&D activities carried out by the global alliance of the Consultative Group on International Agricultural Research (CGIAR) within its agricultural research institutions Centro Internacional de Agricultura Tropical (CIAT) and Centro Internacional de Mejoramiento de Maíz y Trigo (CIMMYT) have been in the process of developing cassava and maize as well as other crops biofortified with pro-V A. These cultivars are expected to reach their target levels, and hence, be ready for introduction in the year of 2009 (personal communication, H. Pachón) to countries, in which a high cost- effectiveness and a high economic impact are expected, on the basis of ex-ante impact assessments. 1.3.1 Literature search A literature search was performed using PubMed, Web of Science and GoogleScholar (September 2008). The search included literature in Danish, English and Spanish using the below-listed keywords in different cross combinations of the Boolean operator AND. Subsequently, the abstracts were read in order to eliminate articles with information not relevant to this thesis. In addition, the reference lists of the selected articles were skimmed to identify other publications of relevance. Websites of AgroSalud, CGIAR, HarvestPlus, the United Nations Children’s Fund (UNICEF), the World Health Organization (WHO) and CIMMYT were also accessed and screened. Key words Vitamin A, β-carotene, β-cryptoxanthin, dietary intake, xerophthalmia, mortality, measles, maize, cassava, Disability Adjusted Life Years, biofortification, fortification, supplementation, cost-effectiveness, bioavailability, bioefficacy, Colombia, Honduras, Mexico, Nicaragua 3 2 The Magnitude of Vitamin A Deficiency (VAD) V AD is a state of excessive depletion of the body’s V A stores, followed by failure to maintain serum retinol concentrations and hence support crucial functions of V A in peripheral tissues. V AD was first linked with disorders of the vision, tissue differentiation and immune function during the 1920s to 1950s following the discovery of V A, in 1913 (Ross 2006). Furthermore, during the 1980s the first relationships between V AD and mortality were established (Sommer et al. 1983; Sommer et al. 1986) and since then V AD has received a great deal of attention due to the magnitude and severity of these adverse outcomes on human health. 2.1 The burden of VAD Recently, due to their high burden of disease, several issues of malnutrition, including V AD, have been identified among the top priorities to be addressed in developing countries in order to advance global welfare. Among these and other major global challenges, supplementation of V A and zinc have been ranked as highest priority based on their perceived costs and benefits (Copenhagen Consensus 2008). Thus, in developing countries currently suffering from a high incidence of V AD, this condition may constitute a serious constraint to human and economic development. 2.1.1 Global status Each year an estimated 107.4 million pregnant women suffer from V AD. 6.2 million of these women eventually develop maternal night blindness (XN), of which over 60% occur in South and Southeast Asia (West 2002). Recent estimates based on the DALYs approach showed that V AD is also currently considered to be responsible for 5.3% of the global childhood years lost to disability, and 0.6 million deaths per year worldwide (Black et al. 2008). The majority of this burden appears to be confined to South Central 4 Asia and several sub-regions of Africa, however, widespread V AD in children <5 years of age also appears in some parts of Latin America and the Caribbean (Figure 1). Figure 1. Prevalence of V AD in children <5 years of age. Adapted from Black et al. 2008. Among the world’s children <5 years of age suffering from V AD, an estimated 6.5% of them live in the Americas (West 2002). 2.1.2 Status in Latin America and the Caribbean V AD was first discussed as a public health problem in Latin America after the worldwide survey conducted by the WHO in 1962 (Chopra & Kevany 1970). However, in the most recent review concerning this region, V AD was considered mostly subclinical and of varying public health importance (Mora, Gueri, & Mora 1998). Nevertheless, each year an estimated 8.2 million preschool children and 375,000 women in the reproductive age suffer from V AD in the Americas, representing 17.3% and 3.8% of these population groups, respectively, while X affects 20,000 preschool children and 376,000 women (West 2002). 2.1.3 Status in Colombia, Honduras, Mexico and Nicaragua Among the four countries included in this thesis, V AD appears to constitute a serious public health problem among children in Mexico only (Table 1), by definition of a prevalence of V AD >15% in children <5 years of age (Sommer & Davidson 2002). 5 Nicaragua has virtually eradicated V AD in preschool children affecting <2%, while the extent of V AD in neighboring Honduras is on the borderline of public importance. The prevalence of maternal V AD, as defined by the International Vitamin A Consultancy Group (IV ACG) (Ramakrishnan & Darnton-Hill 2002), indicated a health problem of public importance in all countries of investigation except Mexico, which was on the borderline. Thus, none of the countries has yet succeeded in eradicating V AD in women. Table 1. Country prevalence of V AD and estimated absolute number affected in Colombia, Honduras, Mexico and Nicaragua. V AD (<0,7 μmol/L) % (a) Number of people affected Children Pregnant Lactating (b) (b) (c) Country <5 women women Children <5 Pregnant women Lactating women Colombia 5.9† 15.3† 15.3† 261,842 135,252 131,330§ Honduras 13.0† 5.3 5.3 122,590 10,547 10,125 Mexico 23.0† 4.2 4.2 3,342,400 88,578 68,205 Nicaragua 1.8Ψ 5.3 5.3 20,801 7,367 69,99 (a) Data derived from West (2002) (b) Data derived from population numbers from UNICEF (2007) (c) Data of percentage exclusively breastfed were obtained from WHO (2008b) and applied to (b) †Prevalence was derived directly from the Vitamin and Mineral Nutrition Information System (WHO 2008a) §The prevalence of breastfeeding, defined as “ever breastfed” was derived from ICBF (2006) ΨPrevalence was derived from Gurdian et al. (2005) In terms of clinical manifestations of V AD, secondary data on the prevalence of X in children <5 years of age is unavailable. However, the presence of maternal X, did not indicate that V AD should be regarded a public health problem in reproductive women in any of the countries, as defined by a prevalence >5% (Sommer & Davidson 2002), although Honduras is on the borderline (Appendix A). 2.2 Major VAD disabilities and outcomes The major outcomes arising due to V AD are those of X and mortality, resulting from a sequelae of disabilities which may predispose to or arise from low V A-status. 2.2.1 Xerophthalmia (X) X is a condition, which involves a series of different ocular abnormalities (Table 2) ranging from the reversible condition of XN to irreversible childhood blindness (Sommer & Davidson 2002; WHO 1996). 6 Table 2. X classification according to severity. Classification Diagnosis/definition VA treatment status XN Nightblindness X1A Conjunctival xerosis Reversible X1B Bitot's spots X2 Corneal xerosis X3A Corneal ulceration/keratomalacia (involving <1/3 of the corneal area) X3B Corneal ulceration/keratomalacia (involving >1/3 of the corneal area) Irreversible XS Corneal scar (from X3) XF Xerophthalmic fundus Modified from the publication of Sommer & Davidson (2002). Evidence originating from case control studies in developing countries severely affected with V AD suggests that X is positively associated with insufficient dietary intake of V A (Shankar et al. 1996; Tarwotjo et al. 1982). Furthermore, the identification of retinal photoreceptors in the eye lends support to the biological rationale that inadequate provision of V A can cause XN. These receptors synthesize rhodopsin from retinal, an aldehyde of V A, which initiates signal transduction of light when illumination is low. In individuals suffering from V AD, rhodopsin levels may decrease due to decreased serum retinol levels, which in turn decrease the responsiveness to light signals. X, until the stage known as corneal xerosis, is considered reversible if sufficient V A is provided (Sommer & West 1996). Ocular responsiveness to V A-treatment of XN (Congdon et al. 1995) and Bitot’s spots (X1B) (Sovani et al. 1994) seems to be enhanced when baseline serum retinol levels are low, which demonstrates a dose-response utilization of V A. However, V AS studies of ocular recovery in humans indicate some interaction with protein-energy malnutrition (PEM) (Brown, Gaffar, & Alamgir 1979; Sommer & Tarwotjo 1982), and, more recently, with deficiencies of riboflavin and iron (Graham et al. 2007), which may be exerted through their effects on serum retinol. 7 2.2.2 Mortality Mortality is multifactorial in nature, especially in developing countries where multiple deficiencies and lack of healthcare persist. It has been proposed that 3% of mortality in children <5 years of age is caused by V AD (Jones et al. 2003) and that it is a consequence of X, infections, measles, and blindness (Sommer & West 1996) in the context depicted in Figure 2. Infections Mortality Low dietary supply of VA Vitamin A deficiency X3 Measles Other nutritional deficiencies XN X1B XS Blindness Figure 2. Immediate factors contributing to the development of V AD, and the relation between major functional outcomes of this condition. Modified from the unpublished study of Rastogi & Mathers 2002. With respect to the role of X, immunologic disturbances may precede or arise during this condition, which may render the subjects more prone to detrimental infections, and hence, mortality (Sommer, Katz, & Tarwotjo 1984; Sommer, Tarwotjo, & Katz 1987). An almost linear positive relationship between mild X, defined as XN and B1X, and mortality has been reported by Sommer et al. (1983), suggesting that X may account for or precede part of the V AD associated mortality. With respect to severe cases of X, half of the children normally die within a year of going blind (Sommer & West 1996). Whereas diarrhoea, is considered an established preventable infection, the actions of V AS are more established for treatment than for prevention of measles (Jones et al. 2003). 8 3 Causes of VAD The causes of V AD can be divided into immediate causes or underlying causes due to the nature of the contexts. While immediate causes are directly related with the body metabolism of V A, underlying causes are highly social and biological. 3.1 Immediate causes The etiology of V AD is multifactorial with immediate causes ranging from nutritional disorders, which involve either inadequate intake or absorption of V A or interacting nutrients, as well as health conditions involving increased excretion of V A. Dietary provision of V A to the human body is essential, due to the body’s inability to synthesize retinol. The liver comprises the primary storage and contains >90% of total body V A, which serves to maintain serum retinol homeostasis. Mobilization of hepatic V A stores occurs during times of inadequate V A supply, in order to meet the needs of peripheral body tissues. If this process continues over a longer period of time the liver may become depleted and hence, unable to sustain serum retinol concentrations within a normal physiologic range of 1.5 – 3.0 μmol/L (Ross 2006). However, several nutrients other than V A have also been proposed to modulate the kinetics of V A (Christian et al. 2001a; Dijkhuizen et al. 2004; Graham et al. 2007; Munoz et al. 2000; Wieringa et al. 2003), which in turn, also appear to be affected by V A status (Allen et al. 2000; Robles-Sardin et al. 1998). Furthermore, V A excretion, if highly increased, as in periods of fever and infections with measles or parasites, can cause low serum retinol (Jalal et al. 1998; WHO 1996). The biologically most vulnerable groups are those in whom the V A requirement is elevated due to increased cell proliferative growth, such as growing children and infants, and pregnant and lactating women. 3.2 Underlying causes Underlying causes of V AD appear to be deeply rooted in social and political issues with poverty, unhealthy environment, and low maternal education being among the major interacting and reinforcing factors, predisposing to the development of V AD (UNICEF 9 1997; WHO 1996). As reflected in the first MDG (SCN 2004), problems of malnutrition and poverty are highly intertwined and should be targeted simultaneously (Figure 3). Poverty Unhealthy Environment Household Food Insecurity Inadequate access to health services Undernutrition Infection Low Maternal Education Own and Family Health Child Feeding Practices Figure 3. Underlying factors contributing to the development of undernutrition and infection. Inspired by UNICEF 1998. Household food insecurity is the main contributor to become underweight and develop malnutrition disorders of PEM and micronutrient deficiencies (UNICEF 1997). Furthermore, V A-rich foods, especially those of animal origin are often less available to the poor, meaning that they rely on fruits and vegetables as their main sources of V A (section 4.3.1) (WHO 1996). These V A sources may be highly seasonal in rural areas due to underdeveloped food systems and need to be consumed in large amounts in order to maintain V A status due to lower bioavailability of carotenoids. Maternal education is essential to family health as women are the principal caretakers of children, and has proved an important positive predictor of serum retinol in children in Bangladesh, whereas infection proved a negative predictor (Kongsbak et al. 2006). The magnitude and the contribution of the reinforcing components shown in Figure 3, may vary in a country specific manner. Of the Latin American countries under investigation, clustering of the above presented manifestations would be expected, due to the presence of political and economic structures causing large disparities notorious for Latin America (Albalak et al. 2000; PAHO 2007b). Colombia, Honduras, Mexico and Nicaragua (Figure 4) have yet to 10 achieve the high standards of living present in affluent countries as evidenced from the indicators below, including per capita Gross National Product (GNP), underweight and literacy (Table 3). Table 3. Indicators of developmental and health character in Colombia, Honduras, Mexico and Nicaragua. Per capita Prevalence of GNP (b) underweight* Adult literacy <5 (b) (a) Country USD (%) (%) Colombia 1,200 7 92.4 Honduras 2,740 11 80.0 Mexico 7,870 5 90.5 † Nicaragua 1,000 10 81.3 Ψ (a) Data were derived from PAHO (2007b) (b) Data were obtained from UNICEF (2007) † Only reported for women, 2000; Ψ population > 10 years *Defined as moderate or severe underweight Figure 4. Overview of Latin America and the Caribbean with the four countries under investiga- tion. Downloaded and modified September 2008 from http://vivirlatino.com/i/2008/01/LANDING- MAP-LATIN-AMERICA.jpg, Despite these problems, the region of Latin America and the Caribbean is expected to achieve MDG one, reducing by half the proportion of the population suffering from extreme hunger and the population living in extreme poverty, defined as living on less than one dollar a day, by the year of 2015 (Pinstrup-Andersen 2007). 3.3 Assessment of VAD Hepatic V A concentration is considered the “Gold Standard” for assessment of V A status, however, it is inappropriate to measure in humans due to the invasiveness of this procedure. Thus, identification of V A deficient individuals in large study populations, instead applies a set of established biochemical cut-offs, as depletion of hepatic V A <20 μg/g will cause disruption in serum retinol homeostasis, although serum levels may not drop immediately (Sommer & Davidson 2002). However, serum retinol levels <0.7 μmol/L (20 μg/dL) are considered deficient and continue to be an important indicator of V AD, especially in young children, when ocular manifestations of X are absent. For 11 lactating women, breast milk retinol concentrations <20μg/dL also serve as a reliable biochemical reference in large scale assessment of V AD, although a history of XN during a past pregnancy resulting in a live birth (<3 years ago) is a well established assessment procedure. Currently V AD is considered to constitute a health problem when >15% of the children and women sampled demonstrate V A deficient serum retinol levels (Sommer & Davidson 2002), and when >5% of reproductive women report having had XN during a recent pregnancy (Ramakrishnan & Darnton-Hill 2002). Furthermore, supportive indicators, such as mortality rates (Schultink 2002), dietary patterns, including breast- feeding, infectious disease prevalence, literacy and income levels, may aid to identify communities, regions, or nations, in which V AD is prevalent (WHO 1996). Among the biochemical assessments largely applied for smaller scale interventions with V A are the modified relative dose-response (MRDR) and the relative dose-response (RDR) tests, which utilize one and two blood tests, respectively, to indirectly measure hepatic V A (Underwood 2004). These tests are considered more responsive to interventions with V A than serum retinol, thus better reflecting the hepatic reserves (Tanumihardjo 2004). However, the deuterated retinol isotope dilution (DRD) test can detect changes in total body reserves of V A. Other applied measures of short-term evaluation include responses in dark adaptation by means of the pupillary threshold (PT), which is mostly applied to women of reproductive age suffering from XN at baseline (Congdon et al. 2000; Graham et al. 2007; Haskell et al. 2005). PT has also been applied to young children <5 years of age with promising results of utilization in the future (Congdon et al. 1995). In Colombia, Honduras, Mexico and Nicaragua, diet records suggest that intake of V A is inadequate in certain population groups as determined by the estimated average requirement (EAR), which is 210 to 275 µg RAE/day for children aged 1-5 years and 485 to 500 µg RAE/day for non-pregnant women of reproductive age (IOM 2001). The EARs for pregnant and lactating women range from 530 to 550 µg RAE/day and 1200 to 1300 µg RAE/woman/day, respectively. According to the reported numbers presented below (Figure 5), the EARs for non-pregnant and pregnant women of the United States and Canada (IOM 2001) appear to be reached by Colombia and Nicaragua only, while none of the countries seem to reach the EARs for lactating women. 12 Figure 5. Current dietary vitamin A consumption of women of reproductive age and preschool children in Colombia, Hondu- ras, Mexico and Nicaragua. Sources are listed in Appendix B. For children from 1-5 years of age, the EARs appear to be met in all countries of investigation with the exception of Honduras. Thus, current V A supply from dietary sources alone appears inadequate to cover the most basic needs of some of the population groups in the countries of investigation. 13 VA intake (µg/RAE/day) 4 Strategies to eliminate VAD The major strategies of addressing V AD include V AS, commercial fortification with V A and dietary diversification. The sustainability of these strategies seems to be inversely related with efficacy, with highly efficacious programs of V AS and fortification often being affected by political instability, international funding, and the lack of evaluation (Darnton-Hill & Nalubola 2002; Underwood 2004). Nevertheless, dietary diversification, being considered more sustainable, is generally less cost-effective, due to the relatively low efficacy of carotenoids and the high costs of nutrition education (Phillips et al. 1996). 4.1 Vitamin A (VA) supplementation V AS, which involves either orally administered or injected V A, is the most conventional form of intervention in the prevention and treatment of V AD in high-risk populations. Large scale studies have reported decreases in the incidence or prevalence of X (Djunaedi et al. 1988; Katz et al. 1995; Sinha & Bang 1976; Vijayaraghavan et al. 1984) and in the incidence of mortality (Sommer et al. 1986) in preschool children after receiving a high dosage of 200,000 International Unit (IU) V A at least every six months. Effects of V AS on mortality outcomes, including other dosing regimens, have largely been confined to children aged 6 – 71 months (Rahmathullah et al. 1990; West et al. 1991), although some studies have reported effects of supplementing infants shortly after birth (Rahmathullah et al. 2003) and women during pregnancy and post-partum (Christian et al. 2000). It is currently recommended by the WHO, that V AS involves administration of a high dose of V A (200,000 IU) at least every six months (WHO 1996), as decreases in serum retinol or deterioration in pupillary scores have been observed hereafter (Congdon et al. 1995). With respect to V A treatment of measles, a recent review revealed an effect of two doses of V AS only in children <2 years, reducing the risk of mortality and pneumonia-specific mortality, which are often secondary to measles (Huiming, Chaomin, & Meng 2005). V AS is currently considered the most cost-effective V AD eradication strategy worldwide (Copenhagen Consensus 2008), although secondary to fortification within some countries 14 of Latin America (Phillips et al. 1996), where coverage rates are generally reasonably good. Of the countries investigated, Honduras, Mexico, and Nicaragua were reported to provide V AS to children <5 years of age, with the highest single-dose coverage rate observed for Nicaragua (Table 4). Table 4. Current V AD strategies and their coverage in Colombia, Honduras, Mexico and Nicaragua. Vitamin A supplementation Fortification (% coverage 2006) (a) (b) At least one dose Full dose Sugar Margarine Milk (IU V A/kg) (IU V A/kg) (IU VA/kg) Colombia - - 30,000 Honduras 40 40 50,000 35,000 2,000 Mexico 68 63 20,000 1,033-2,033 Nicaragua 98 0 > 16,600 (a) Data obtained from UNICEF (2007) (b) Data obtained from Darnton-Hill & Nalubola (2002); personal communication, H. Cori Although having a lower coverage rate, Honduras and Mexico, seem to provide full dose to the percentage of the population reached, whereas no information was available for Colombia. 4.2 Fortification of commercial foods with VA Fortification is defined as adding a micronutrient to a food vehicle whether the nutrient is already found in the food vehicle or not. Poor government control of fortified products has at times resulted in markedly different levels of fortification than those considered mandatory (Darnton-Hill & Nalubola 2002; Ribaya-Mercado et al. 2004). Furthermore, fortified products have been poorly evaluated in terms of their effectiveness. In a small scale study carried out in Nepal, consumption of V A-fortified rice for 60 days produced significantly lower serum retinol and less improvement in PT scores among women with baseline XN than did those consuming liver (Haskell et al. 2005). However, in Guatemala, after one year of consuming V A-fortified of sugar 76% of the children investigated experienced an increase in serum retinol (Arroyave, Mejia, & Aguilar 1981), and fortified sugar was later demonstrated to provide 25% alone of the V A-intake 15 received from the complementary diet of urban toddlers (Krause, Delisle, & Solomons 1998). However, fortification has yet to prove as effective in rural areas as demonstrated in urban areas (Nestel et al. 1999) due to lack of infrastructure (Pinstrup-Andersen 2007). Current programs of V A fortification in Colombia, Honduras, Mexico and Nicaragua utilize margarine, milk and sugar as vehicles (Table 4). However, among these countries, studies evaluating the effects of these mandatory fortification programs remain limited. One evaluation applying the DRD test, carried out in Nicaragua showed that the current levels of sugar-fortification significantly improved mean serum retinol, hepatic V A concentrations; and total body V A reserves in 21 children aged 5-9 years after one year of mandatory fortification of sugar with V A (Ribaya-Mercado et al. 2004). These V A stores were similar to those reported for children in USA at follow-up, with some exhibiting V A stores >1.05 µmol/g liver. 4.3 Dietary diversification The aim of dietary diversification strategies is to increase production, intake and bioavailability of natural food sources of V A, conventionally achieved through education of communities (Faber et al. 2002). There is evidence emerging in support of a future role for food-based strategies in preventing V AD in humans. Studies have shown ability of an increased consumption of V A-rich fruits or vegetables to maintain or improve the V A status as evidenced by RDR and DRD in high-risk populations after a period of only 20- 70 days of consumption (Haskell et al. 2004; Mariath, Lima, & Santos 1989; Ncube et al. 2001; Ribaya-Mercado et al. 2007; Tang et al. 1999). However, the efficacy of this strategy is considered limited due to the type of V A derived from plant sources, and the effects of preparation on carotenoid bioavailability, that is the fraction of pro-V A absorbed and available for bioconversion. 4.3.1 Natural food sources of VA The Institute of Medicine (IOM) defines V A as a generic term, which includes the fat- soluble compound retinol (Figure 6), its aldehydes, and all pro-V A carotenoids from 16 which retinol can be synthesized in the human body (2001). Retinol is found in animal products, whereas natural sources of pro-V A carotenoids are leafy green vegetables, yellow fruits and tubers. Upon ingestion both compounds are solubilized into micelles in the intestinal lumen for cellular uptake, however, retinol is more efficiently absorbed and stored in the body than pro-V A due to differences in actions of absorption (Ross 2006). The unit retinol activity equivalent (RAE) has been introduced to reflect carotenoid bioconversion to retinol from non-specific dietary sources, which is predominantly carried out in the small intestinal mucosa. 4.3.2 Carotenoid bioconversion Although there is knowledge of more than 600 carotenoids, only β-carotene, α-carotene and β-cryptoxanthin (Figure 6) appear to be important sources of V A, of which β- carotene is considered the most potent in terms of V A activity (IOM 2001; Ross 2006). Figure 6. Chemical structure of all-trans-retinol, all-trans-β-carotene, all-trans -α-carotene and all-trans-β-cryptoxanthin. Adapted from IOM (2001). An established bioconversion relationship built on the unit of RAE states that two µg of all-trans-β-carotene, when dissolved in oil, can yield one µg of all-trans-retinol, whereas other V A active carotenoids are considered half as efficient. However, bioconversion of 17 these carotenoids from all dietary sources are considered 6 times less efficient than when dissolved in oil due to limiting factors related to absorption (IOM 2001). In addition, bioconversion rates also appear to be affected by the nutritional state of the individuals (Ribaya-Mercado et al. 2000). Evidence from rat studies suggests that a possible mechanism for increased bioconversion in the V A-deficient state could be mediated through increased activity (van Vliet et al. 1996; Villard & Bates 1986), rather than synthesis (Gronowska-Senger & Wolf 1970), of β-carotene-15, 15´-dioxygenase, the enzyme that cleave β-carotene to retinal. 4.3.3 Factors affecting the bioefficacy of pro-VA carotenoids The term, bioefficacy, describes the fraction of an ingested nutrient, which is absorbed and converted to its active form. Bioefficacy of dietary pro-V A in the human body appears to be influenced by a series of factors related to compound release of the matrix, and disturbances in micelle formation causing impaired bioavailability (Table 5). Table 5. Major factors affecting bioavailability of dietary carotenoids. Substance/processing Action on bioavailability Action specificity Dietary fat, lipids Promotes Dietary fibre Inhibits Fibretype dependent Disruption of matrix Promotes Heat treatment, puréeing Human studies have demonstrated a superior bioconversion of carotenoids from fruits than vegetables (de Pee et al. 1998; Khan et al. 2007), although bioconversion rates appear to be vegetable specific and highly variable ranging from 10:1 to 27:1 (Haskell et al. 2004; Tang et al. 1999; Tang et al. 2005). Several human studies have demonstrated improvements in bioavailability of pro-V A from vegetables after disruption of the matrix by means of heat treatment or puréeing (Castenmiller et al. 1999; Rock et al. 1998; Yeum & Russell 2002), whereas bioavailability of raw products appear to be highly individual (Edwards et al. 2002). With respect to heat treatment, the transformation from all-trans- β-carotene to various cis-isomers, has demonstrated to produce increased micellarization in in-vitro studies using Caco-2 cells (Ferruzzi et al. 2006; Liu, Glahn, & Liu 2004; Ryan et al. 2008), however, preferential uptake of all-trans-β-carotene has been reported in 18 humans (Gaziano et al. 1995). Dietary fat is also considered important for micelle formation, and it has been suggested that only 3-5 grams may be needed for optimal absorption of synthetic β-carotene (Roodenburg et al. 2000). Nevertheless, dose- responses in chylomicron carotenoid content have been reported with higher fat intakes than those suggested (Brown et al. 2004; Dimitrov et al. 1988; Jalal et al. 1998). In addition, a recent study conducted on gerbils also suggested a significant effect of adding more dietary fat, not only by enhancing carotenoid uptake but also via increasing the bioconversion of these compounds (Deming et al. 2000). In the same study, the soluble fibre, citrus pectin, posed a limiting factor of carotenoid bioavailability and bioconversion when consumed with β-carotene. Human studies have demonstrated different effects of dietary fibre, however, a dependence on the type of fibre has been reported (Richelle et al. 2004; Riedl et al. 1999), with soluble fibres negatively affecting bioavailability. Thus, the restraints on bioavailability may render food-based strategies less efficacious than fortification and V AS, however, they provide an alternative solution less dependent on government funds and initiatives in the long term, once nutritional knowledge has been adopted into common practice. 19 5 Biofortification Biofortification is defined by a genetic improvement in a cultivar’s density of a nutrient. This can be achieved through genetic engineering, however, if natural high-nutrient species of the cultivars in question exist, the developmental practice of cross-breeding can be applied (Nestel et al. 2006). As about three quarters of the world’s poor depend directly or indirectly on agriculture in terms of employment and nutrition (Diaz-Bonilla & Robinson 2001), targeting the base of a food system with biofortification is warranted. 5.1 Rationale for cross-breeding cassava and maize for a higher pro-VA content Although, there is a large presence of dietary sources rich in pro-V A in Latin America, such as pumpkin, sweet potato, palm fruit, buriti, and papaya, most fruits, including the widely consumed mango, are seasonal, and hence not available all year round (Rodriguez-Amaya 1999b). Traditional staple cultivars, on the other hand, have a low content of pro-V A, but are in constant demand and hence continuously consumed throughout the year. Cassava (Manihot esculenta) and maize (Zea mays Sp.), both indigenous staples to the Americas, each possess natural high pro-V A species, and are able to grow in poor soils and demand little maintenance, which make them appropriate for cross-breeding and subsistence agriculture in marginal lands (CGIAR 2008). Furthermore, in Mexico, maize is consumed by the urban poor and people residing in rural areas, contributing 20% and 48% of the energy intake of adolescent women, respectively (Lozada et al. 2007). Thus, indications are that poor people rely on cheap staples for consumption, which may justify their utilization in alternative food-based approaches such as biofortification. 5.2 The beneficiary system of Research & Development (R&D): improving VA status and health outcomes The overall aim of biofortifying staples with pro-V A is to enhance V A status and prevent adverse outcomes of V AD through human consumption of these staples in developing 20 countries affected by V AD. In order to achieve this, biofortified cultivars must be adopted by both farmers and consumers at risk of developing V AD (Figure 7), meaning that cultivar performance, affordability, accessibility and acceptability for the poor are essential factors. However, the stability of carotenoids is very sensitive to light and post- harvest processing, both of which are known to accelerate their degradation (Chávez et al. 2008; Rodriguez-Amaya 1999a). IMPACT New incidence of VAD outcomes = gains in DALYs and USD National govern- mental or non- Adoption governmental efforts and investments Bioefficacy Bioconversion Intake of target cultivars Dose-response Post-harvest VA-status Nutrient density Dissemination of cultivars biofortified with pro-VA Investments of AgroSalud R&D Figure 7. Technical beneficiary cycle of biofortification, and factors affecting the potential health and monetary impacts. Preliminary studies indicate that post-harvest retention of pro-V A in these cultivars may prove an additional challenge with reported levels varying between 17.3% and 71.1% for 21 cassava after different methods of handling and preparation (Chávez et al. 2008). Because post-harvest handling is determined culturally by how food is traditionally prepared this constraint may prove difficult to modify in order to achieve a higher health impact of the biofortified cultivars. 5.3 Bioefficacy of biofortified staple cultivars Bioavailability and bioconversion of the pro-V A carotenoids present in the biofortified cultivars comprise part of the concerns of bioefficacy. Until now, studies investigating the efficacy of the dominant carotenoids present in cassava or maize, whether occurring in natural or enhanced amounts corresponding to those attained through biofortification, are limited to animal studies applying Mongolian gerbils. 5.3.1 Gerbil studies of biofortified components dissolved in oil The dominant carotenoids found in enhanced amounts of the pro-V A-biofortified cassava and maize under investigation are β-carotene and β-cryptoxanthin, respectively (personal communication, N. Palacios). To date, there are only a few studies reporting on bioefficacy of pro-V A-biofortified cultivars, including those of maize and carrots, which were consumed as powder dissolved in oil by V A-depleted Mongolian gerbils. These studies confirm that bioconversion of the carotenoids present in the pro-V A-biofortified maize appear to be at least as effective as provision of supplemental β-carotene (Davis et al. 2008a; Davis et al. 2008b). The supplemental β-carotene did, however, appear less efficient (4.6-2.52 µg β-carotene to 1 µg retinol) than suggested by the IOM, which is in line with other gerbil studies receiving β-carotene supplements dissolved in oil (Howe & Tanumihardjo 2006; Tanumihardjo & Howe 2005). With respect to hepatic V A repletion, several studies have demonstrated that the pro-V A carotenoids present in the biofortified products are bioavailable from maize and carrots and sufficiently efficacious to replete and maintain hepatic V A stores (Davis et al. 2008a; Davis et al. 2008b; Dosti et al. 2006; Howe & Tanumihardjo 2006; Sulaeman et al. 2002). Surprisingly, the ability of β- cryptoxanthin from biofortified maize to raise hepatic V A has proven enhanced to that of supplemental β-carotene in one study (Davis et al. 2008a). A superior effect over β- 22 carotene supplementation was also evident after consumption of the powder of a β- carotene-biofortified carrot, although the increase in hepatic V A appeared less than would be expected, indicating decreased bioconversion when higher amounts of β- carotene are consumed (Dosti et al. 2006). However, improvements in hepatic V A positively correlated with the amount of β-carotene received from a β-carotene- biofortified maize in another study (Howe & Tanumihardjo 2006). Other studies, however, indicate that dietary provision of β-carotene, appears to be only as efficacious as supplemental β-carotene, which in terms of efficacy approximates half that of supplemental V A (Howe & Tanumihardjo 2006; Sulaeman et al. 2002). 5.3.2 Human consumption studies of pro-VA-biofortified cultivars To date, there are only two studies investigating the effects of human consumption of cultivars biofortified with pro-V A solely, of which one is a controlled efficacy study and the other a community trial. The pro-V A-biofortified cultivar examined in both studies was orange-fleshed sweet potato (OFSP). Daily consumption of OFSP supplying 1031 RAE of beta-carotene for 53 days resulted in significantly improved hepatic V A, by means of the MRDR test, in intervention children aged 5-10 years in relation to the control group (van Jaarsveld et al. 2005). Furthermore, the proportion of children with adequate hepatic V A stores tended to increase in the intervention children, whereas the proportion with low serum retinol decreased significantly in both groups. However, in another study, two years after the introduction of OFSP in two districts of Mozambique the prevalence of serum retinol levels <0.7 µmol/L was significantly lower in intervention children of 2-5 years of age compared with controls (Low et al. 2007). This study, which involved market development efforts of farmer extension coupled with nutrition education increased the consumption of OFSP in intervention children, which consumed OFSP significantly more frequently than did controls. Thus, there appears to be potential for pro-V A-biofortified cultivars, to increase V A status in humans. However, more human studies are needed to elucidate the efficacy and acceptability of pro-V A- biofortified cassava and maize in prospective countries of introduction. 23 6 Methodologies 6.1 Conceptual framework of the Disability Adjusted Life Years (DALYs) formula The health and economic impact of introducing pro-V A-biofortified cultivars were calculated using the DALYs method, which applies the prevalence of established functional outcomes of V AD to assess the amount of years of life lost to disability. The years of life lost to disability include those lost to short-term illness (YLDtemp), permanent disability (YLDperm) and mortality (YLL) (Murray & Lopez 1996). DALYslost = YLL + YLDperm + YLDtemp In the modified formula below, only mortality and disability are compiled separately: 1− e−rL j  1− e−rdij  DALYslost = ∑ T M   j j   + ∑ ∑ T j I ij D   j r i j ij      r  Tj = Total number of people in the target group j Mj = Rate of mortality associated with the deficiency in the target group j Lj = Average life expectancy remaining in target group j Iij = Rate of incidence of illness i in target group j Dij = Weight of the incapacity due to illness i in target group j dij = Duration of illness i in the target group j r = Rate of discount for the years of future life Note: In the case of the parameter dij, permanent diseases are equal to the average of the remaining life expectancy Lj. The DALYs model applied had been specifically constructed by HarvestPlus to include dietary aspects of improvement, information on disability weights, illness duration, and rate of discount in relation to V AD outcomes (Stein et al. 2005). This model disregarded the age-weighting in the generation of income and applied national Standard Life 24 Expectancy (SLE) which were among the major modifications made in relation to the original formula (Murray 1996). Therefore, average per capita income was applied to represent the monetary value per DALY gained, while the costs per DALY represented those of R&D related to adaptive breeding and subsequent dissemination, however, these were not country specific in this analysis. 6.2 Selection criteria of target countries Initially, data collection was attempted for 13 countries from the region comprised by Latin America and the Caribbean, including Bolivia, Brazil, Colombia, Costa Rica, Cuba, Dominican Republic, El Salvador, Guatemala, Haiti, Honduras, Mexico, Nicaragua and Peru. However, the criteria of providing secondary dietary data representing intake of V A and the respective cultivars, for women of reproductive age and children <5 years separately (Figure 8), was met by Colombia, Honduras, Mexico and Nicaragua only (Appendix B). Children <5 years Women of reproductive age Demographic data Demographic data Data on pregnant population: % live births Data on % live births: pregnant population Unavailable for all countries Unavailable for all countries lactating population Data on lactating population: Unavailable for all countries Health data Data on XN, XS, blindness, XN, XS, blindness measles with. complications: Health data XN available for: measles, unavailable for all countries. XN Bolivia, Dominican Republic, measles with Measles available for: Honduras, Peru complications all countries VA intake available for: VA intake available for: Nutritional data Colombia, Honduras, Mexico, Nutritional data Honduras, Mexico, Nicaragua VA intake, Nicaragua VA intake, Cassava intake available for: intake of cassava Cassava-intake available for: intake of cassava Colombia, Honduras, Nicaragua intake of maize Colombia, Honduras, Nicaragua intake of maize Maize intake available for: Maize intake available for: Colombia, Honduras, Mexico, Colombia, Honduras, Mexico, Nicaragua Nicaragua Figure 8. Flow diagram on the availability of collected secondary data critical to the DALYs model. (Not all data needed to conduct the analysis are included). 25 6.2.1 Population and demographic data Data of demographics were obtained from major international organizations, including UNICEF and the WHO. Whenever available, 2006 was the reference year of the population and demographic data. Population estimates of children <5 years of age and pregnant women in the respective countries were obtained from the UNICEF State of the World’s Children Report 2008, in which 2006 was the most recent reference year (UNICEF 2007). The number of pregnant women, because not readily available in the statistics, was assumed to reflect the annual number of births reported. The population size of lactating women was calculated by multiplying the number of births with the percentage of children ever breastfed as reported from the most recent Demographic and Health Survey (DHS) obtained from the WHO website (WHO 2008b) and in the case of Colombia from Instituto Colombiano de Bienestar Familiar (ICBF 2006). The number of live births, and perinatal deaths from the year 2000 were obtained from the report on Neonatal and Perinatal Mortality (WHO 2006) to calculate the percentage of live births as outcome of all pregnancies:  perinatal deaths  %live births = 1−    (live births + perinatal deaths)  Furthermore, SLE at birth, per capita GNP, crude birth rate (CBR), under 5 mortality rate (U5MR) and, solely for country comparison purposes, moderate and severe undernutrition, wasting, and stunting were obtained (UNICEF 2007). 6.2.2 Health data Prevalence of measles in children <5 years of age was obtained from the Immunization Newsletter published by the Pan American Health Organization (PAHO 2007a). Furthermore, national experts were consulted. National prevalence of X for women was obtained directly from the Vitamin and Mineral Nutrition Information System (VMNIS) only if surveys were reported after 2001 or not providing prevalence on a national level (WHO 2008a). When the most recent information of the prevalence of X was provided 26 before 2001 the data utilized for the calculations were extracted from West (2002) (Appendix C) as these estimates were made nationally representative and weighted for future impacts of V A program coverage initiated at the time of data collection. Prevalence of childhood X was not reported by VMNIS and estimates by West (2002) were reflected by the absence of studies. Data was therefore obtained from expert consultation or extrapolations made from other countries within and outside the Latin American region. The comparison analysis was performed with selected countries showing relatively similar profiles of the prevalence and severity of V AD as mapped by West (2002). Among these countries, only those which had readily available national data on childhood X obtained from the VMNIS where matched with the countries under investigation according to U5MR ranging within the following category 20-<40. Subsequently, the matching countries were selected based on the similarity of their profile in the prevalence of moderate and severe stunting, underweight, or wasting as defined by UNICEF. A deviance of up to 25% in any one of these variables applied for comparison was accepted as a cut-off. Furthermore, the human development index was applied for comparison whenever needed (UNDP 2008). On the basis of the data selected to estimate childhood X, the prevalence of corneal scars (XS) was calculated applying a ratio observed between XN plus X1B and XS on a regional level for Latin America and the Caribbean in 1990 (Rastogi & Mathers 2002). The prevalence of XS was calculated in the following manner, as X1B was included in the definition of X reported by West (2002): XS (%) 0.02  X (%)⋅100 = ⋅     102  In this way, the value of XS was adjusted to the value of X, which was then calculated by subtracting the prevalence of XS. Due to lack of information of the prevalence of keratomalacia (X3), the prevalence of XS alone served to calculate the prevalence of childhood blindness, by multiplying by a factor of 0.5 (Stein et al. 2005) (Appendix A). 27 6.2.3 Nutritional data National nutrition surveys reporting individual level median consumption of V A in the respective population groups within the last ten years from the reference period 2006 were included. The definition of nationally representative data applied in this thesis was when the word ‘national’ appeared in the title of articles recovered from a database search performed on PubMed or Web of Science, or when data were defined as national by the respective organizations providing the data. In general, consumption data obtained from women of reproductive age (preferably 15-49 years) were applied to represent those of pregnant and lactating women, in the absence of biological distinctions in the data available. For the countries where the best available national level nutritional data were provided by the World Bank’s Living Standard Measurement Survey (LSMS) or the International Food Policy Research Institute (IFPRI), food composition tables from the Food and Agriculture Organization (FAO) had been applied by non-nutritionist HarvestPlus staff to derive the mean amount of V A consumed by women > 14 and children ≤ 5 years of age. However, because these surveys represented the household level intake, statistical multiple regression analysis had been performed on the data set in order to obtain the separate intakes of individuals of the household. In addition, data from the LSMS (World Bank 2001) and the Family Allowance Program (PRAF) (IFPRI 2003) also served to provide a habitual mean consumption of cassava and maize. If not readily available, these data were derived with the help of statisticians on accessible national datasets deriving median consumption, or the Food Balance Sheets (FBS) available at FAOs website (FAO 2003) (Appendix C). 6.3 Target cultivars assumptions According to the DALYs methodology, optimistic and pessimistic scenarios were estimated with regards to expected final pro-V A content, post-harvest losses, bioconversion, and technology coverage rate for the respective crops, while investments related to their development and introduction were quantified for one scenario only. 28 6.3.1 Optimistic and pessimistic scenarios Optimistic and pessimistic assumptions were made on the basis of single available studies, consultancy with expert personnel from the breeding centres or extrapolation from other ex-ante publications (Appendix D) (Stein et al. 2005; personal communication, J. V . Meenakshi). 6.3.2 Financial investments Financial investments related to R&D, strengthening of National Agricultural Research Systems (NARS), varietal development, program management, administration, seed systems and dissemination were extracted solely from current investments of AgroSalud that initiated in 2004 and continuing to 2009. Because these were sunk costs, discounting was not applied to these investments, during the cost-effectiveness calculations of short- term character. An economist from AgroSalud had calculated the average of these costs in relation to expected beneficiary countries for the respective crops, disregarding the size of the population of the countries involved (personal communication, S. S. Pérez). 6.4 Sensitivity analysis performance Sensitivity analyses were carried out separately for the four most questionable data applied in the calculations, including childhood X, V A-intake of women of reproductive age in Colombia, bioconversion rates of cassava and maize and income levels. Comparisons were obtained from applying plausible values, versus the values applied, in the DALYs calculations. In relation to the factor, income levels, values previously applied to define economic gains by HarvestPlus served to reflect the comparison for the analysis (Meenakshi 2007). Sensitivity was then evaluated by means of absolute and relative terms of the change in either one or more of the DALYs results, including DALYs gained, cost-adjusted monetary gain, Internal Rates of Return (IRR) and Benefit/Cost (B/C), depending on the factor investigated. 29 7 Ex-ante impact results for Colombia, Honduras, Mexico and Nicaragua 7.1 Current burden of VAD The calculated absolute loss of DALYs currently attributed to V AD per year without the introduction of pro-V A-biofortified crops, varied greatly within the four countries of investigation, incurring annual health costs ranging from USD 4.7 million in Nicaragua to USD 509.8 million in Mexico (Table 6). Table 6. DALYs lost to V AD per year in Colombia, Honduras, Mexico and Nicaragua. Loss Colombia Honduras Mexico Nicaragua Due to mortality in children <5 15,772 4,491 64,774 4,303 Due to temporal disability 3,220 999 0 338 Due to permanent disability 150 78 0 56 Total due to disability 3,370 1,077 0 394 Total 19,142 5,568 64,774 4,697 In monetary terms (million USD) 52.4 6.7 509.8 4.7 Disability, including maternal and childhood X, appeared to account for a smaller part of the DALYs lost due to V AD, except in Mexico, where it constituted no burden. However, DALYs lost to mortality among preschool children seemed to comprise, by far, the largest burden of V AD in all countries. 7.2 Potential health and economic impact of biofortication with pro-VA 7.2.1 DALYs gained due to biofortified cassava The calculated health gains of introducing pro-V A-biofortified cassava were highest in Nicaragua (Table 7), where the calculated efficacy levels reached 100% for preschool children (Appendix E). Thus, if provided to all Nicaraguan children in current average amounts recorded, 97.5% of the children will receive enough V A, as exemplified by reaching the recommended dietary allowance (RDA) for V A in this population group. Nicaragua appeared to be the only country in which this staple is consumed in sufficient 30 amounts to prove economically viable on a short-term basis, as evidenced by positive monetary gains after the total costs were incurred in both scenarios. However, in the remaining countries, introducing the biofortified cassava can prove economically viable on a short-term basis only if the optimistic scenario will prevail, with cost-adjusted annual monetary gains ranging from USD 0.2 million in Colombia to USD 1.6 million in Nicaragua. Table 7. Gains in DALYs, USD and reduction of the health burden with biofortified cassava per year. Country Scenario DALYs gained Monetary gain Health burden reduction (years) (million USD) (%) Colombia Optimistic 147 0.2 0.8 Pessimistic 7 -0.2 0.0 Honduras Optimistic 400 0.3 7.2 Pessimistic 20 -0.2 0.4 Mexico Optimistic 140 0.9 0.2 Pessimistic 7 -0.1 0.0 Nicaragua Optimistic 1,815 1.6 38.6 Pessimistic 250 0.1 5.3 The expected health burden reduction in the optimistic scenario was 38.6% in Nicaragua, but appeared rather negligible in the other countries ranging from 0.2% in Mexico to 7.2% in Honduras. The main beneficiaries of introduction of this food-based strategy seemed to be preschool children through averting V AD related mortality, except in Colombia, where the children were recorded to have a higher V A intake than the RDA before the introduction of any biofortified cultivars. Thus, according to the DALYs model, there was no response in Colombian children to the consumption of either biofortified cassava or maize. 7.2.2 DALYs gained due to biofortified maize According to the calculations, the only country in which introducing pro-V A-biofortified maize did not appear economically viable on a short-term basis at present was Colombia in the pessimistic scenario, due to a potential negative cost-adjusted monetary gain (Table 8). In the other countries, these annual gains ranged from USD 2.2 million in Nicaragua to USD 80.9 million in Mexico in the optimistic scenario, with monetary gains ranging from USD 0.6 to 9.8 million, respectively, as evidenced by the results of the pessimistic 31 scenario. Health burden reductions were highest in Nicaragua, reaching 49.2% and, when excluding Colombia, lowest in Mexico only reaching 15.9% in the optimistic scenario; however, the highest gains of DALYs were recorded in Mexico. Table 8. Gains in DALYs, USD and reduction of the health burden with biofortified maize per year. Country Scenario DALYs gained Monetary gain Health burden reduction (years) (million USD) (%) Colombia Optimistic 259 0.6 1.4 Pessimistic 31 -0.1 0.2 Honduras Optimistic 2,386 2.7 42.8 Pessimistic 453 0.4 8.1 Mexico Optimistic 10,291 80.9 15.9 Pessimistic 1,268 9.8 2.0 Nicaragua Optimistic 2,309 2.2 49.2 Pessimistic 761 0.6 16.2 In general, health gains, as reflected by gains in DALYs and reductions in the health burden, were more pronounced with the biofortified maize than cassava, which could be ascribed to a higher expected pro-V A content and a higher recorded habitual intake of maize in the countries. In Nicaragua, consumption of these two cultivars appeared to have an age component, meaning that children <5 years of age were recorded to consume more cassava than maize, whereas the women of reproductive age had a higher recorded intake of maize than cassava. 7.2.3 New incidence of functional outcomes of VAD Reductions XN, but not in XS, childhood blindness, measles or measles with complications were to be expected from introducing any of the pro-V A-biofortified cultivars under investigation in any of the countries as evidenced by the DALYs calculations (Appendix F). Reductions in childhood XN up to 39% and 49% for Nicaragua were expected to be attained in the optimistic scenario from introducing cassava and maize, respectively, whereas reductions in maternal XN were 18% and 45% for pregnant women, and considerably lower for lactating women (data not shown). Apart from Colombia, reductions in the health burden were predominantly confined to averting mortality of children <5 years of age, which seemed to constitute a major part of the V AD related annual loss of DALYs in all countries. Up to a 50% reduction in child 32 mortality appeared to be expected from introducing biofortified maize in Nicaragua, which seemed to achieve the highest increment in overall health burden reduction (Table 9). Overall, biofortified maize appeared to be more effective than cassava in reducing the current burden of mortality of children <5 years of age. Although a higher absolute number of childhood deaths were averted in Mexico, this country did not achieve a high relative reduction in comparison to Honduras and Nicaragua. This became more evident in the presentation of the mortality results per 1,000,000 children <5 years, or total population in each of the four countries of investigation (Table 10). Table 9. Averted mortality in children <5 years of age per year due to biofortified cassava or maize. Cassava Maize New Mortality No. of Deaths New Mortality No. of Deaths Country Scenario due to V AD deaths averted per due to V AD deaths averted per (% per 1,000 ) from V AD year (% per 1,000 ) from V AD year Colombia Optimistic 3.0 533 0 3.0 533 0 Pessimistic 3.0 533 0 3.0 533 0 Honduras Optimistic 2.8 142 12 1.7 85 69 Pessimistic 3.0 153 1 2.7 140 13 Mexico Optimistic 3.0 2,160 5 2.5 1,821 344 Pessimistic 3.0 2,164 0 2.9 2,122 42 Nicaragua Optimistic 1.8 88 58 1.5 73 73 Pessimistic 2.8 138 8 2.5 121 24 Table 10. Number of deaths averted in children <5 years of age per 1,000,000 due to biofortified cassava or maize. Number of deaths in children <5 years averted per 1,000,000 Among children <5 years Country Scenario Among total population Cassava Maize Cassava Maize Colombia Optimistic 0 0 0 0 Pessimistic 0 0 0 0 Honduras Optimistic 13 73 2 10 Pessimistic 1 14 0 2 Mexico Optimistic 0 33 0 3 Pessimistic 0 4 0 0 Nicaragua Optimistic 86 109 10 13 Pessimistic 12 36 1 4 As evidenced by the new incidence rates (Table 9) and absolute number of childhood deaths averted in relation to the population <5 years or total population (Table 10), both of the biofortified cultivars appeared far more effective in reducing this health burden in Nicaragua and Honduras, than in Mexico. Thus, the most economically viable solutions 33 on short-term (Table 7 & Table 8) were not necessarily the ones saving most DALYs. As seen from the results for biofortified cassava, with Colombia and Honduras each gaining more DALYs than Mexico, these countries did not have a larger economic gain than Mexico due to the difference in per capita GNP (Table 3). 7.3 Potential cost-effectiveness of biofortified cassava and maize The expected short-term cost-effectiveness is determined by a comparison of the costs in the optimistic and pessimistic scenario with the gains of each DALY saved during one year. Introducing pro-V A-biofortified cassava or maize, would achieve the highest cost- effectiveness in Nicaragua and Mexico, respectively (Table 11). However, the costs of USD 110.8 per DALY saved if introducing biofortified cassava in Nicaragua, by far exceeded the costs per DALY saved if biofotified maize would be introduced in Honduras, Mexico or Nicaragua, where the costs were only USD 60.5; 13.6; and 62.5, respectively. The GNP per capita, reflecting the monetary gain of each DALY saved, exceeded the costs in all the optimistic scenarios for both of the biofortified cultivars. However, in the pessimistic scenarios, introducing cassava only appeared short-term cost- effective in Nicaragua, while introducing maize appeared cost-effective on a short-term basis for Mexico, Honduras and Nicaragua. Table 11. Cost and gain in USD per DALY saved by biofortified cassava and maize. Country Gain per DALY Cost per DALY with cassava Cost per DALY with maize Optimistic Pessimistic Optimistic Pessimistic Colombia 2,740 1,335* 28,028 541* 4,521 Honduras 1,200 490* 9,810 59* 309* Mexico 7,870 1,401* 28,028 14* 111* Nicaragua 1,000 108* 785* 61* 184* *Introduction appears to be cost-effective by WHO standards (WHO 2001) While the cost-effectiveness results merely reflect the highest gains expected during one year, IRR and the B/C presented below (Table 12) reflect the gains by means of the discounted future earnings over a 20-year period in relation to the incurred total costs. These results show that introducing any of the two biofortified cultivars could be 34 profitable in all countries on a long-term basis, with the exception of cassava if the pessimistic scenario would prevail in Colombia, which presents an IRR of –1%. Table 12. IRR and B/C over a 20 year period for biofortified cassava and maize. Cassava Maize Country Scenario IRR (%) B/C IRR (%) B/C Colombia Optimistic 43 14 54 25 Pessimistic -1 1 17 3 Honduras Optimistic 46 17 86 101 Pessimistic 1 1 49 19 Mexico Optimistic 71 55 216 4,005 Pessimistic 15 3 133 494 Nicaragua Optimistic 75 64 81 82 Pessimistic 34 9 55 27 Overall, in terms of IRR and B/C, introducing biofortified maize appears more effective than cassava ranging from low values of 17% and 3 in Colombia with the pessimistic scenario to high values of 216% and 4,005 in Mexico, respectively. 7.4 Sensitivity to selected parameters of the DALYs model 7.4.1 Childhood X estimations The sensitivity of the results demonstrating the current burden of V AD, when adjusting for the childhood X extrapolations applied to Colombia, Honduras and Nicaragua, as compared to the scenario of a zero prevalence, appeared to be rather low, merely ranging between a 1.1% to 2.2% decrease in total DALYs lost and a 1.5% to 2.2% decrease in the monetary loss (Appendix G) of the results presented (Table 6). This factor appeared to induce some decreases in the IRR of introducing cassava and maize, which did not prove significant for the outcome of long-term cost-effectiveness in Nicaragua (data not shown). For Honduras, however, the long-term outcomes of IRR and B/C in the pessimistic scenario of introducing cassava no longer proved cost-effective if there would be no childhood X. 35 7.4.2 Bioconversion of beta-carotene and beta-cryptoxanthin The sensitivity of bioconversion applied in the calculations that led to the results presented above (section 7.2 & 7.3) proved very country specific, but overall insignificant for the outcome of short-term economic viability and the long-term outcomes of IRR and B/C, although the values for these parameters did change accordingly (Appendix H). In the optimistic scenario for cassava and maize a higher bioconversion rate of 6:1 was compared to that of 12:1, which was the one applied to obtain all results in all scenarios presented in this section. Large differences were observed in short-term economic gains, which in the case of cassava tripled in Colombia while only doubled in Honduras and Mexico. In Nicaragua the gains remained similar, due to only minor improvements in the DALYs saved through decreasing maternal XN. For maize, only minor increments in short-term economic gains were observed in Honduras while these almost doubled in Colombia and Mexico, with no effect in Nicaragua. In addition, both scenarios were compared with the higher bioconversion rate of 24:1 for maize, instead of the overall applied 12:1, which did not appear to have any significant consequence for IRR, B/C or short-term economic viability in any of the scenarios, again with negligible effect of these parameters in Nicaragua with exception of the pessimistic scenario. 7.4.3 Income levels of target population The sensitivity of income levels in terms of short-term and long-term cost-effectiveness was relatively high and caused several changes of significance if introducing cassava only when an arbitrary value of USD 1,000 was applied to reflect the gain per DALY saved in Colombia, Honduras and Mexico (Appendix I). These results changed from positive to negative values for IRR in the pessimistic calculations carried out with cassava and B/C ratios ranging from 0.3 to 0.7 in Honduras and Mexico, respectively (data not shown). Furthermore, it would no longer be considered cost-effective on a short-term basis to introduce cassava in the optimistic scenario in Colombia and Mexico. The appliance of a lower arbitrary value of USD 500 in Nicaragua caused a change in short-term cost-effectiveness of cassava as evidenced by a negative gain (data not shown), however, this income value did not appear to cause any significant changes in the viability outcome of IRR and B/C for cassava or maize. 36 7.4.4 Dietary VA intake in Colombia Sensitivity to the reported dietary intake of V A in Colombia, which was estimated as the mean intake of the reported intake of women aged 14-18 and 51-64 years, was very low and did not yield any changes in monetary gains of introducing biofortified cassava reported to an accuracy of USD 0.1 million (Appendix J). Only minor changes in the DALYs saved, IRR, and the health burden reduction appeared to constitute the sensitivity of this factor on the outcome, within the range applied. 37 8 Discussion Eight ex-ante assessments were carried out to determine the V AD alleviating potential of introducing pro-V A-biofortified cassava or maize in Colombia, Honduras, Mexico or Nicaragua in two scenarios. The DALYs calculations indicated that a considerable annual health and, hence, economic loss due to V AD could be expected in all four countries. The major contributor to the loss of DALYs appeared to be childhood mortality, rather than maternal or childhood X. Decrease in mortality of children <5 years of age also appeared to constitute the majority of the DALYs saved with the biofortified cultivars, in three of the four countries. However, there are issues concerning the general validity of the data applied in the DALYs calculations, and the strengths and limitations of the analytical framework used, which must be taken into consideration. 8.1 Validity and influence of data applied in the DALYs calculations The parameters presented in the following sub-sections were those considered subject to either large variance or uncertainty. In order to test their influential value on the results presented (section 7.1; 7.2; 7.3), their sensitivity analyses are discussed below. 8.1.1 Estimations of childhood X In receiving expert consultancy from Mexico, it was revealed that there had been no observed cases of severe V AD or X in the most recent study conducted in 2006 (personal communication, J. R. Dommarco). No expert information was received from the other countries and, therefore extrapolations were made. These were carried out in accordance with an unpublished study of WHO applying the U5MR for comparison purposes (Rastogi & Mathers 2002). However, the U5MR is affected by a number of nutritional and health problems and therefore does not solely reflect V AD, and a rate <50 may not even reflect the presence of V AD (Schultink 2002). This was the case with all countries under investigation. It means that in spite of emerging evidence that, at least in Nicaragua, V AD is no longer a major public health problem (Gurdian et al. 2005), the 38 extrapolations suggest the presence of X. Secondly, West (2002) defined X in the countries of comparison as a compilation of XN, X1B, X3 and XS, which led to XS being estimated by means of a regional factor reported for XN and XS (Rastogi & Mathers 2002). The sensitivity results showed that if these countries would not present any cases of X, these extrapolations would not appear to affect the estimation of the current burden of V AD by more than a decrease of 2.2% in monetary losses calculated for Nicaragua (Appendix G). This indicates that the burden of childhood X, whether obtained from estimations or country expert consultations, may exert little influence on the V AD burden in the countries under investigation. However, it can affect the question of long-term cost-effectiveness of introducing biofortified cultivars, when predicted short-term economic gains are negligible, as found for cassava in the pessimistic scenario in Honduras. 8.1.2 Dietary intake of VA A number of uncertainties were related to the reported dietary intake of V A, due to differences in data recollection and agency funding. The LSMS from Nicaragua (World Bank 2001) and the nutritional survey in Honduras (IFPRI 2003) both reported dietary intake based on household consumption during the preceding fifteen and seven days, respectively; whereas the nationally conducted nutrition surveys in Colombia and Mexico were carried out on an individual level by obtaining 24-hour dietary recalls (Appendix E). The 24-hour recalls, although considered biased due to underreporting in Mexico by the authors (Barquera et al. 2003a), provided the most reliable nutritional data applied in these ex-ante analyses. However, an obvious bias from applying the national survey data from Colombia was that it did not present dietary intake of V A for women of reproductive age. Nevertheless, data were available for other population groups stratified by gender and age, and served the purpose of performing a sensitivity analysis. A very narrow range of ± 10 µg RAE/woman/day for the estimated intake was applied for the sensitivity analysis as it constituted the range of difference in the dietary data available for the younger and older women (Appendix B). This may explain the negligible differences observed when changing the value for this parameter (Appendix J). It is, however, not known whether the actual intake would fall within this range. Nevertheless, 39 a supporting factor giving rise to these speculations was that the intake of energy exhibited a similar pattern to that applied for estimating V A intake (ICBF 2006), although a poor correlation has been demonstrated between energy intake and carotenoids with the 24-hour recall method (Bingham & Day 1997). Furthermore, the household surveys from Honduras and Nicaragua did not appear to account for the way in which the food was prepared. However, indications of post- harvest processing were available for maize, reported as whole or milled, but not for cassava. In addition, it could be speculated that the fortified products currently on the market and part of the habitual diet were not reported separately or distinctly as fortified in these two countries. This poses a risk that these foods were simply converted to conventional nutrient values through use of the traditional food composition tables from FAO. Thus, it is likely that the contribution from fortification was unaccounted for. Both countries did have mandatory fortification of sugar with V A at the time of dietary collection, which in neighbouring Guatemala was shown to constitute 25% of the total V A intake alone from non-breast milk food sources of children aged 6-36 months (Krause, Delisle, & Solomons 1998). When adding fortified Incaparina and margarine to the equation, the share of V A obtained from fortified foods by these toddlers increased to 90%. In Honduras, the reported V A intake of preschoolers and their mothers appeared extremely low in relation to the prevalence of V AD. Although being hit by hurricane Mitch just prior to the dietary recollection, the presence of additional mandatory V A fortification of margarine and milk could play a large role in the supply of V A, although the exact coverage is unknown. In addition to the possible lack of reporting consumption of fortified foods, it could be speculated that consumption of additional V A-rich foods than those appearing in the dietary questionnaires, applied in Honduras and Nicaragua, were left unreported by the population questioned. Moreover, the widely applied strategy of government V AS is complementary to all the diet records obtained. 8.1.3 Bioconversion rates for the carotenoids in cassava and maize Bioconversion, until human efficacy studies are carried out with cassava and maize, remains a matter of uncertainty, due to the conflicting evidence of current guidelines and 40 preliminary studies with gerbils. The different carotenoids present in biofortified maize, in particular, should in theory yield a lower bioconversion rate than that (12:1) advised to be applied in the calculations (personal communication, N. Palacios). This is due to the large presence of β-cryptoxanthin, which is considered half as effective as β-carotene in terms of bioconversion (IOM 2001). Furthermore, processing, preparation, and V A status, all factors difficult to make adjustments for, appear to be closely related to bioavailability, and hence bioconversion. Preliminary studies carried out in gerbils indicated that the natural dosing regimen of providing β-carotene or β-cryptoxanthin as part of a diet can yield higher bioconversion rates, than those observed for supplemental synthetic β-carotene when dissolved in oil (Davis et al. 2008a; Davis et al. 2008b). This may be explained by a dependency of passive diffusion in relation to carotenoid uptake (Riedl et al. 1999), which did not appear to constitute a constraint for β-cryptoxanthin when supplied to gerbils in twice the amount than β-carotene (Davis et al. 2008a). However, although gerbils exhibit similarities to humans in carotenoid metabolism (Lee et al. 1998; Thatcher, Lee, & Erdman 1998) these studies should be viewed with scepticism and call for further investigation. Gerbils appear to have a different lipid profile from that in humans, and as carotenoids are incorporated in lipid for transportation (Dosti et al. 2006), serum β-carotene and bioconversion may be affected differently in gerbils and humans. Nevertheless, the validity of the current guideline of 24:1 for β- cryptoxanthin (IOM 2001) may be questioned. This rate was used for comparison in a sensitivity analysis for maize only. As a bioconversion rate of 6:1 was applied in the optimistic scenario of another DALYs publication (Zimmermann & Qaim 2004), this rate was compared for both cassava and maize in another sensitivity analysis. This did not result in major changes with respect to short- or long-term economic viability in this analysis. 8.1.4 Income levels of the target populations The monetary value of a DALY saved differs according to the reporting organization ranging from USD 200 to anywhere between the single to triple value of the GNP reported by the World Bank and the WHO, respectively (Stein, Sachdev, & Qaim 2006; 41 WHO 2001). However, due to the high association of poverty and malnutrition coupled with the large disparities in income and standard of living typical for Latin America (PAHO 2007b), the population groups deemed to benefit most from these pro-V A- biofortified cultivars will have an income below the average GNP. The sensitivity analysis revealed that a change in income levels could induce major changes in short- and long-term economic viability, in countries with high per capita GNP values, and small gains in DALYs. In this analysis, changes in short-term economic viability were observed for both scenarios when investigating the potential of cassava in Colombia and Mexico. 8.2 Understanding and interpreting the results from the DALYs model The modified DALYs formula provides a simplified framework for presenting the potential health and economic benefits of biofortification to governments and NGOs. Among the noticeable challenges with the application of this model in countries with subclinical V AD was the lack of essential data and the detachment of V A intake and V AD related outcomes. For instance, in Colombia, mortality of children <5 years of age was predicted by the DALYs formula to comprise a relatively large burden at baseline due to the assumption that 3% of childhood mortality is caused by V AD (Jones et al. 2003). Nevertheless, the introduction of any of the pro-V A-biofortified products was predicted to have no effect on this or any of the other parameters related to childhood V AD, because current reported intake of preschool children exceeded the RDA. This may explain the calculated negligible gain in DALYs observed for this country, as the majority of the DALYs saved in the other three countries were gained from averting deaths of children <5 years of age. Furthermore, with respect to childhood mortality, the health impact of biofortification was assumed to avert only two of the 3% V AD related deaths. Although measles is no longer a problem in Latin America (PAHO 2007a), the remaining 1% of V AD related mortality was considered related to measles in the DALYs model, and increasing the dietary content of pro-V A does not cure measles (Jones et al. 2003). Hence, reduction rather than eradication of V AD was the only possible outcome of the model used. 42 8.2.1 Outcomes of current analyses in relation to other DALYs-based assessments Most of the assumptions and values applied in the current ex-ante calculations (Appendix E) were similar to those applied by Zimmermann & Qaim (2004) in the first ex-ante assessment of introducing pro-V A-biofortified rice in the Philippines. They reported a total gain of 15,311 - 85,137 healthy life years, representing an economic gain in the range of USD 15.8 - 87.7 million, without adjusting for any costs. In comparison, the highest gains in DALYs emerging from the present analysis were 1,815 in Nicaragua and 10,291 in Mexico for cassava and maize, respectively. However, potential monetary gains similar to those in the optimistic scenario of the Phillipines were achieved if introducing maize in Mexico only, due to a considerably higher per capita GNP. In Latin America, the potential of introducing pro-V A-biofortified cultivars measured in DALYs has been estimated on a regional level for cassava in the V AD affected Northeast Brazil only (Meenakshi et al. 2007), which exhibited a V AD burden similar to that of Mexico in absolute numbers calculated. In the calculations for Northeast Brazil, a health burden reduction of 19% in the optimistic scenario was to be expected with costs per DALY saved ranging from USD 126.50 to 1,006.46. Despite a lower absolute gain in DALYs, a cost of only USD 13.6 per DALY saved was expected in this thesis for maize in Mexico, which was predicted a health reduction of 16%. However, a value of only USD 1,000 was applied to reflect the monetary gains in Northeast Brazil (Meenakshi et al. 2007), and the costs of biofortification additionally included costs of nutrition education and maintenance breeding. These costs were expected to amount to USD 3 million over a 30 year period rather than the USD 0.20 million expected over a 20-year period used in this thesis for all four countries. This may explain the relatively large difference in the short- term cost-effectiveness of Northeast Brazil and Mexico. 8.2.2 Comparative costs of biofortification in relation to current VAD strategies Presently, in relation to other strategies for combating V AD, biofortification of cassava or maize with pro-V A did not appear to provide more short-term cost-effective than current V AD strategies, or provide a solution to the V AD related problems faced by the four 43 countries in question. According to Phillips et al. (1996), investigating the cost- effectiveness of V AD preventative programs carried out in neighbouring Guatemala, the most cost-effective strategy identified was fortification with a cost of USD 0.98 per high risk person reached. In the present analysis, saving one DALY with the introduction of pro-V A-biofortified maize in the optimistic scenario in Mexico would incur a cost of USD 13.6. This by far exceeds the annual costs of the most expensive intervention identified by Phillips et al. (1996), food production and education, ranging between USD 3.10 – 4.16 per high risk person reached. However, it should be noted that the costs applied by Phillips et al. (1996), although negligible, did not include preparatory costs, whereas the costs of biofortification are almost exclusively those of R&D and dissemination efforts. These costs are considered a one time investment (Nestel et al. 2006), as opposed to the costs reported for V AS and fortification on an annual basis. Thus, long-term results should be emphasized for this food-based strategy, which achieved an IRR and a B/C of 216% and 4,005, respectively, for introducing biofortified maize in Mexico. Importantly, the strategy of introducing either of the two biofortified cultars appeared to be cost-effective on a long-term basis in all four countries under investigation, except Colombia, if the pessimistic scenario will prevail with biofortified cassava. 8.2.3 National or regional approach: effect of aggregate intake data It can be argued that a more targeted ex-ante analysis may be needed, especially in the two larger countries, due to variances in V AD prevalence, and consumption of V A, cassava and maize. A regional rather than national introduction of biofortified cultivars may prove more effective in terms of lowering the costs per DALY saved and increasing the IRR. However, in countries with subclinical V AD, investigations and reports of V AD are generally lacking, also regionally. However, available data of reported median intakes of the present cultivars in Colombia and Mexico indicate that consumption is highly skewed, also regionally. Indications are that the biofortified cultivars may benefit a smaller part of the population than that assumed. Coverage rates of consumption among the target groups were expected to range between 15-40% and 25-50% for cassava and maize, respectively; however, in Colombia only 19.3% of the total population questioned 44 in the national nutrition survey (ICBF 2006) reported to consume cassava regularly, amounting to an average of 83.2 g/person/day. It was also reported that maize was consumed only by 6.2% with an average ingestion of 87.5 g/person/day, while maize flour was consumed also by 6.2% with an average intake of 49.7 g/person/day. Unfortunately, this consumption information was not reported by region, or stratified by age, gender or socioeconomic status (SES) making it difficult to identify, in where consumption levels of the target groups would be high enough to have a positive impact on V A status. Whether the two cultivars reportedly consumed were own produce or commercially acquired was not accounted for in any of the countries included in this analysis, however, it may influence the health impact considerably. Moreover, due to lack of other sources, the daily average intake of cassava in Mexico was obtained from national statistics on commercial agriculture of import and export, showing the amount of cassava available per capita per year (FAOSTAT, 2003). This source may not only underestimate the actual intake in developing countries, due to the large presence of subsistence agriculture, but may also largely fail to reflect the food habits of poor farmers not engaging in commercial activities. This is, however, important, as large inequities exist with respect to V A consumption in Mexico, where rural areas have a median adequacy of V A intake of 56.1%, as opposed to 98.9% in urban areas (Barquera et al. 2003b) demonstrating the scale of national heterogeneity. Due to these concerns, a recent impact study conducted by Stein et al. (2006) applied a different methodology, which utilized disaggregated individual reported intakes from national surveys to account for this limitation. Furthermore, health data, SES and policy factors were also applied, and the cost of each DALY saved with golden rice in V AD affected India would be as low as USD 3.1 (Stein, Sachdev, & Qaim 2006). In essence, it would probably improve cost- effectiveness of the countries under investigation if the analysis was approached on a regional level and if representative individual level dietary data was applied disaggregated, especially in Colombia and Mexico. 8.2.4 The beneficial scope of the DALYs results An important aspect of the DALYs model, is the extensive beneficial potential of biofortification if introduced and commonly consumed by part of the women and children 45 specifically targeted in this ex-ante analysis. This includes the hidden assumption that improving maternal V A status through increased V A consumption transfers more V A to the lactating infant, considering that all children <5 years of age are included in this analysis. However, studies of V AS and β-carotene supplementation are unequivocal with respect to this mother-infant dyad. There is reason to believe that V A status of the mother plays a crucial influence in childhood mortality as maternal XN during pregnancy has been associated with low-birth weight and an increased risk of infant morbidity in India (Tielsch et al. 2008). While mortality of infants with xerophthalmic mothers appeared to be significantly reduced by maternal V AS, this effect seemed to be less pronounced with β-carotene supplementation, administered from six months before pregnancy (Christian et al. 2001b), and another study failed to produce any effect with a weekly low dose of either supplementations (Katz et al. 2000). However, other studies on post-partum maternal V AS and β-carotene supplementation suggest beneficial effects on infants younger than six months of age through increasing the V A content in breast milk, as a result of improving V A status of the mother (Rice et al. 1999; Stoltzfus et al. 1993). Whether the improvements achieved in V A status will reflect those predicted by the ex- ante assessments, is difficult to estimate, and may also depend on breast-feeding practices. A direct limitation in the DALYs model, however may be that certain less equivocal nutritional benefits are left unaccounted for, such as the effects of improved maternal survival attained with V AS and β-carotene supplementation during pregnancy (West et al. 1999), on the survival of infants or young children. Furthermore, the remaining population groups may also experience improvements in V A- and health status from consuming the biofortified cultivars, which is among the benefits exceeding the scope of the DALYs results presented in this thesis. 8.3 Factors that may influence the achievement of the calculated impacts Through consumption of the pro-V A-biofortified cassava or maize, expected health and economic impacts within the range of the pessimistic to optimistic scenario should be achieved. Among the factors determining the levels of achievement with these 46 biofortified cultivars are seed dissemination, adoption, food preparation, cultural beliefs, and the co-occurrence and underlying causes of deficiencies of other nutrients. 8.3.1 Political commitment to ensure adoption The efficiency of the biofortified cultivars under investigation appear to be highly dependent on political involvement of national governments, partly due to the need of government financing of social marketing in addition to the investments of AgroSalud. As political instability is a continuous problem in Latin American countries and one of the main problems of current V AD strategies (Darnton-Hill & Nalubola 2002), adoption and, hence, coverage rate may be adversely affected if dissemination efforts rely on national funding. In comparison with fortification, biofortified cultivars face the additional challenges of not only having to be adopted by the consumer, but also by the producer (Nestel et al. 2006). This may present a problem as breeders often use different criteria for evaluation of cultivars than consumers. Adding additional qualities to the biofortified cultivars in order to increase yield, which is part of the breeding process financed by AgroSalud, would be important for the acceptance by farmers. However, a study on maize acceptability in Oaxaca, Mexico also demonstrated a concern on the farmer level for taste and suitability for preparation of special dishes (Bellon et al. 1998). The change in colour, an inevitable part of increasing the pro-V A content in these cultivars may also affect acceptance and hence consumption. Five out of nine varieties of OFSP were accepted for cultivation after introduction through a community based field intervention study in Mozambique, which after two years demonstrated a significantly increased cultivation and human consumption substituting the white sweet potato commonly consumed in that area (Low et al. 2007). A home garden project in South Africa also demonstrated acceptance of OFSP in areas where this cultivar had not been available with 19% of the intervention children consuming OFSP at least once a week 20 months after implementation (Faber et al. 2002). Whether similar results can be achieved without comprehensive farmer extension and nutritional education, which characterized the two latter studies, needs to be elucidated. 47 In order to maximize adoption, dissemination of the biofortified cultivars as seeds or produce could be for free, as an integrated part of federal health projects to minimize dissemination costs. The health project “Oportunidades” in Mexico, targets poor people with similar characteristics of the ones included in this ex-ante analysis, and provides them with money and fortified foods in exchange for showing up to medical checks and attending school (Rivera & Sepulveda 2003), and there is a similar project, PRAF, in Honduras, which could also serve the purpose of dissemination (IFPRI 2003). The idea with biofortification of targeting the poor people cultivating and consuming their own produce is to complement current efforts by overcoming the constraint of poverty. However, if a premium price could be obtained for the biofortified products, those farmers may favour the increased earnings over consumption of their produce, enabling them to buy more food, possibly including better sources of V A for their families. Thus, farmer adoption does not ensure consumer adoption, however, it may lead to increased V A intake either way if the demand is high. 8.3.2 Culture, food habits and efficacy Natural high pro-V A varieties of the biofortified cultivars under investigation exist, however, cultivation of these varieties, have been abandoned, probably due to specific desirable properties of the white maize currently grown (personal communication, N. Palacios). In addition, post-harvest processing inevitably depends on cultural practices related to preparation and consumption of the target crops, which may affect bioefficacy (de Pee et al. 1995) and, hence, impact on V A status of those consuming the cultivars. Although appearing as a separate strategy for preventing V AD, education in food production and preparation could be regarded integral elements of investment in the implementation and adoption of biofortification as it has been in other DALYs assessments for the same cultivars (Meenakshi et al. 2007). Government or NGO led programs may be indispensible in changing barriers of cultural beliefs and practices. An important aspect is child feeding as complementary food in developing countries is typically low in lipids, which may result in lower carotenoid bioavailability. Although, it has been reported that Mexican children aged 12-23 months receive 32% of their energy intake from lipids, far lower proportions have been reported for other Latin American 48 countries (Lutter & Rivera 2003). Parental education seems to be crucial in caring for their offspring in many developing countries as evidenced with respect to the OFSP intervention in Mozambique. Education increased nutritional knowledge in both men and women in the intervention area, and their children aged 2-5 years demonstrated significantly higher serum retinol levels than children from the control areas, probably due to the significantly increased intake of OFSP (Low et al. 2007). However, changing food habits requires investments into nutrition education programmes, which are costs not accounted for in these ex-ante analyses. 8.3.3 Co-presence of other nutritional deficiencies In all four countries of investigation, deficiencies of some interacting nutrients are present and may affect the beneficial effects of consuming an increased amount of pro-V A if they co-occur with V AD. However, whether one nutrient deficiency contributes to the development of the other is difficult to elucidate. Results have been unequivocal in Mexico, with significant improvements observed in serum ferritin levels, reflecting improved iron status, occurring three months after a single dose of V AS (Robles-Sardin et al. 1998), whereas supplementation with iron and zinc provided separately or in conjunction significantly raised serum retinol (Munoz et al. 2000). Co-occurrence of iron deficiency and V AD has also been reported in Honduras in children 1-5 years of age (Albalak et al. 2000), and in Nicaragua it was shown that only a small fraction of reported anemia was caused by iron deficiency in children <5 years of age and women of reproductive age (Gurdian et al. 2005), suggesting a role of V AD or other nutrient deficiencies. This suggests that adequate provision with V A may not solely eliminate V AD and its health outcomes, in the dose-response manner applied in the DALYs model. For instance, β-carotene, only when supplemented in conjunction with zinc to marginally deficient women during pregnancy, increased both maternal and infant serum retinol significantly in comparison with a control group six months postpartum (Dijkhuizen et al. 2004). The infants of the zinc and β-carotene supplemented mothers had an odds ratio of 0.22 compared with infants of the controls for developing V AD at the age of 6 months. The higher plasma retinol levels in the mothers appeared to be mediated by increased cleavage of β-carotene with the administration of zinc. Iron also appeared to have an 49 effect on the distribution of V A, causing significantly decreased serum retinol in Indonesian infants suffering from marginal V AD at baseline after six months of supplementation, however, V A stores simultaneously increased, whereas supplemental β- carotene had no effect on either outcome (Wieringa et al., 2003). On the other hand, improving V A status may contribute to the health benefits brought about by increasing iron or zinc status, and hence, resulting in a larger impact. In essence, the efficacy of an intervention, may be determined the magnitude of certain nutrient deficiencies and health impact achievements may vary according to other health problems of the countries in question. 50 9 Conclusion The current health and economic burden of V AD in Colombia, Honduras, Mexico and Nicaragua is relatively high, especially due to the high U5MRs in these four countries. This burden can be reduced in a long-term, cost-effective manner if pro-V A-biofortified cassava or maize is adopted for cultivation and consumption, without any promotional efforts or further investments. Decrease in deaths of children <5 years of age will constitute the majority of the DALYs saved due to the consumption of either of the two pro-V A-biofortified cultivars in Honduras, Mexico, and Nicaragua. However, in Colombia, due to the methodological limitations, as presented in this thesis, women will be the only beneficiaries, and if the pessimistic scenario prevails, no long-term economic gains are to be expected from investing in breeding efforts and consumption of the biofortified cassava over a 20-year period. In general, biofortified maize will result in higher health and monetary gains in all four countries compared to biofortified cassava. However, in Nicaragua, cassava may prove to be nearly as beneficial, and hence, cost- effective as maize, due to similar levels of consumption. Uncertainties related to the assumptions used in the DALYs model, including those for bioconversion and income, separately resulted in large effects on the health and economic impacts, although these were not crucial for the outcomes. Estimations of childhood X and dietary V A intake of women of reproductive age produced relatively small differences, however, childhood X appeared to modulate the long-term economic outcome of introducing cassava in Honduras only. Thus, current V AD interventions of V AS and fortification should be kept in place throughout the introduction of these biofortified cultivars, due to the uncertainties of the preliminary results presented in this thesis, and the possible need for further investments in nutrition education and social marketing campaigns. 51 10 Future perspectives If adopted by farmers and consumers, the effects of the biofortified cultivars may face external barriers in the future due to a dependency on trade agreements and policies in the global market. An example is the North American Food Trade Agreement, which flooded the Mexican market with imports of cheap maize in the 90s resulting in farmer displacement for those who did not diversify their production away from maize (de Janvry, Sadoulet, & de Anda 1995). Also the relatively low price elasticity for basic staples may result in future substitutions with other less nutrient-dense staples in response to changes in food prices. This could in turn negatively affect consumption of the biofortified cultivars for a period of time, until prices revert to normal levels. Other factors which can be predicted to raise the prices of staple foods include, non-food utilization for example biofuel, which is in progress for maize (Carpita & McCann 2008; Torney et al. 2007) and occurrence of natural disasters resulting in crop failure, which may become more prevalent in the future due to climate changes. In Latin America and the Caribbean obvious countries for the introduction of pro- V A-biofortified varieties include by order of magnitude of V AD in preschool children, Haiti, the Dominican Republic and El Salvador (West 2002; WHO 2008a). Among these countries, El Salvador and Haiti may stand to benefit from maize and cassava, respectively; while the Dominican Republic may benefit more from biofortifying other cultivars. However, in relation to the two cultivars it may benefit the most from cassava which has the highest daily consumption of 35 g/person/day, as reported by the production statistics of FAO (2003). Ex-ante assessments can elucidate the potential health and economic impacts in these countries. Biofortification may only be beneficial at the immediate level of the causes of V AD, and may not contribute to eradicate the underlying causes. However, while achievements of political, social and economic advancements are taking place, biofortification may serve to improve human health, which is a human right. 52 11 Acknowledgements In terms of realization of this thesis, I will like to direct a special thanks to my supervisors in order of degree of supervision Dr. Shakuntala Haraksingh Thilsted, Dr. Per Pinstrup-Andersen and Dr. Henrik Friis for their professional guidance, support and contribution to my work throughout each phase of my work. Also, I wish to thank Dr. Helena Pachón, nutritionist, AgroSalud, CIAT, Colombia, for the collaborative efforts before and during my time there. I would also like to thank her for supporting me at my seminar presentation at CIAT. For daily assistance with the calculations during my stay with AgroSalud at CIAT, I wish to thank Mr. Salomón Suarez Pérez, economist, AgroSalud for always answering my questions with a smile and for taking on the responsibility of collaborating and sharing his work with me. For technical assistance, I will like to thank Dr. J. V . Meenakshi, Impact and Policy Coordinator, HarvestPlus, CGIAR for guiding me through many difficult tasks related to my work with the DALYs model. I am also very thankful to Dr. Keith West, Department of International Health, Johns Hopkins University, USA for providing all his background data on country estimations of V AD from his publication in the Journal of Nutrition, 2002. Furthermore, I wish to thank Ms. Natalia Palacios, microbiologist, CIMMYT, Mexico, for supplying me with information regarding the cultivars. For statistical assistance, I had the pleasure of working with Mr. James Garcia, statistician, CIAT, Colombia after corresponding with Dr. Lars Korsholm, formerly employed at the Southern University of Denmark. I am very grateful for both of their involvement, and wish to thank them for all their help. For supplying country specific information through email correspondence, I wish to thank Dr. Patricia Carillo, nutritionist, CIAT, Nicaragua; Ms. Elpidia Poveda, Ministry of Health, Colombia, Dr. Juan Rivera Dommarco, Director, National Institute of Health, Mexico; and Mr. Héctor Cori, Scientific and Technical Director, Nutrition Improvement Program, Chile. 53 Finally but not least, I wish to thank in order of financial contribution Fællesfonden, PlanDanmark, Knud Højgaards Fond, Oticon Fonden, the Danish International Development Assistance, Ministry of Foreign Affairs of Denmark, the University of Copenhagen, and AgroSalud for their financial contributions to this project. 54 12 References AgroSalud 2008, "The development and deployment of biofortified staple crops to reduce nutrient deficiencies and improve food security in Latin America and the Carribean" [Cited 2008 September]. Available from: http://www.agrosalud.org/index.php?option=com_content&task=view&id=9&Itemid=12. Albalak, R., Ramakrishnan, U., Stein, A. D., Van der, H. F., Haber, M. J., Schroeder, D., & Martorell, R. 2000, "Co-occurrence of nutrition problems in Honduran children", J.Nutr ., vol. 130, no. 9, pp. 2271-2273. Allen, L. H., Rosado, J. L., Casterline, J. E., Lopez, P ., Munoz, E., Garcia, O. P ., & Martinez, H. 2000, "Lack of hemoglobin response to iron supplementation in anemic mexican preschoolers with multiple micronutrient deficiencies", Am.J.Clin.Nutr ., vol. 71, no. 6, pp. 1485-1494. Arroyave, G., Mejia, L. A., & Aguilar, J. R. 1981, "The effect of vitamin A fortification of sugar on the serum vitamin A levels of preschool Guatemalan children: a longitudinal evaluation", Am.J.Clin.Nutr ., vol. 34, no. 1, pp. 41-49. Barquera, S., Rivera, J. A., Espinosa-Montero, J., Safdie, M., Campirano, F., & Monterrubio, E. A. 2003a, "Energy and nutrient consumption in Mexican women 12-49 years of age: analysis of the National Nutrition Survey 1999", Salud Publica Mex., vol. 45 Suppl 4, p. S530-S539. Barquera, S., Rivera, J. A., Safdie, M., Flores, M., Campos-Nonato, I., & Campirano, F. 2003b, "Energy and nutrient intake in preschool and school age Mexican children: National Nutrition Survey 1999", Salud Publica Mex., vol. 45 Suppl 4, p. S540-S550. Bellon, M. R., Smale, M., Aguirre, A., Aragón, F., Suketosh, T., Diaz, J., & Castro, H. 1998, "Maize diversity management and utilization - a farmer-scientist collaborative approach in the central valleys of Oaxaca, Mexico: a progress report.", CIMMYT internal document. Bingham, S. A. & Day, N. E. 1997, "Using biochemical markers to assess the validity of prospective dietary assessment methods and the effect of energy adjustment", Am.J.Clin.Nutr ., vol. 65, no. 4 Suppl, pp. 1130S-1137S. Black, R. E., Allen, L. H., Bhutta, Z. A., Caulfield, L. E., de, O. M., Ezzati, M., Mathers, C., & Rivera, J. 2008, "Maternal and child undernutrition: global and regional exposures and health consequences", Lancet, vol. 371, no. 9608, pp. 243-260. Brown, K. H., Gaffar, A., & Alamgir, S. M. 1979, "Xerophthalmia, protein-calorie malnutrition, and infections in children", J.Pediatr ., vol. 95, no. 4, pp. 651-656. 55 Brown, M. J., Ferruzzi, M. G., Nguyen, M. L., Cooper, D. A., Eldridge, A. L., Schwartz, S. J., & White, W. S. 2004, "Carotenoid bioavailability is higher from salads ingested with full-fat than with fat-reduced salad dressings as measured with electrochemical detection", Am.J.Clin.Nutr ., vol. 80, no. 2, pp. 396-403. Carpita, N. C. & McCann, M. C. 2008, "Maize and sorghum: genetic resources for bioenergy grasses", Trends Plant Sci., vol. 13, no. 8, pp. 415-420. Castenmiller, J. J., West, C. E., Linssen, J. P ., van het Hof, K. H., & V oragen, A. G. 1999, "The food matrix of spinach is a limiting factor in determining the bioavailability of beta-carotene and to a lesser extent of lutein in humans", J.Nutr ., vol. 129, no. 2, pp. 349-355. CGIAR 2008, "Research and Impact: Areas of research: Cassava" [Cited 2008 March]. Available from: http://www.cgiar.org/impact/research/cassava.html. Chávez, A. L., Sánchez, T., Ceballos, H., Rodriguez-Amaya, D. B., Nestel, P ., Tohme, J., & Ishitani, M. 2008, "Retention of carotenoids in cassava roots submitted to different processing methods", unpublished. Chopra, J. G. & Kevany, J. 1970, "Hypovitaminosis A in the Americas", Am.J.Clin.Nutr ., vol. 23, no. 2, pp. 231-241. Christian, P ., Khatry, S. K., Yamini, S., Stallings, R., LeClerq, S. C., Shrestha, S. R., Pradhan, E. K., & West, K. P ., Jr. 2001a, "Zinc supplementation might potentiate the effect of vitamin A in restoring night vision in pregnant Nepalese women", Am.J.Clin.Nutr ., vol. 73, no. 6, pp. 1045-1051. Christian, P ., West, K. P ., Jr., Khatry, S. K., Kimbrough-Pradhan, E., LeClerq, S. C., Katz, J., Shrestha, S. R., Dali, S. M., & Sommer, A. 2000, "Night blindness during pregnancy and subsequent mortality among women in Nepal: effects of vitamin A and beta-carotene supplementation", Am.J.Epidemiol., vol. 152, no. 6, pp. 542-547. Christian, P ., West, K. P ., Jr., Khatry, S. K., LeClerq, S. C., Kimbrough-Pradhan, E., Katz, J., & Shrestha, S. R. 2001b, "Maternal night blindness increases risk of mortality in the first 6 months of life among infants in Nepal", J.Nutr ., vol. 131, no. 5, pp. 1510-1512. Congdon, N., Sommer, A., Severns, M., Humphrey, J., Friedman, D., Clement, L., Wu, L. S., & Natadisastra, G. 1995, "Pupillary and visual thresholds in young children as an index of population vitamin A status", Am.J.Clin.Nutr ., vol. 61, no. 5, pp. 1076-1082. Congdon, N. G., Dreyfuss, M. L., Christian, P ., Navitsky, R. C., Sanchez, A. M., Wu, L. S., Khatry, S. K., Thapa, M. D., Humphrey, J., Hazelwood, D., & West, K. P ., Jr. 2000, "Responsiveness of dark-adaptation threshold to vitamin A and beta-carotene supplementation in pregnant and lactating women in Nepal", Am.J.Clin.Nutr ., vol. 72, no. 4, pp. 1004-1009. 56 Copenhagen Consensus 2008, "Copenhagen Consensus 2008 - results" [Cited 2008 August]. Available from: http://www.copenhagenconsensus.com/Default.aspx?ID=953. Darnton-Hill, I. & Nalubola, R. 2002, "Fortification strategies to meet micronutrient needs: successes and failures", Proc.Nutr .Soc., vol. 61, no. 2, pp. 231-241. Davis, C., Jing, H., Howe, J. A., Rocheford, T., & Tanumihardjo, S. A. 2008a, "beta- Cryptoxanthin from supplements or carotenoid-enhanced maize maintains liver vitamin A in Mongolian gerbils (Meriones unguiculatus) better than or equal to beta-carotene supplements", Br .J.Nutr ., vol. 100, no. 4, pp. 786-793. Davis, C. R., Howe, J. A., Rocheford, T. R., & Tanumihardjo, S. A. 2008b, "The xanthophyll composition of biofortified maize (Zea mays Sp.) does not influence the bioefficacy of provitamin A carotenoids in Mongolian gerbils (Meriones unguiculatus)", J.Agric.Food Chem., vol. 56, no. 15, pp. 6745-6750. de Janvry, A., Sadoulet, E., & de Anda, G. G. 1995, "NAFTA and Mexico's maize producers", World Development, vol. 23, no. 8, pp. 1349-1362. de Pee, S., West, C. E., Muhilal, Karyadi, D., & Hautvast, J. G. 1995, "Lack of improvement in vitamin A status with increased consumption of dark-green leafy vegetables", Lancet, vol. 346, no. 8967, pp. 75-81. de Pee, S., West, C. E., Permaesih, D., Martuti, S., Muhilal, & Hautvast, J. G. 1998, "Orange fruit is more effective than are dark-green, leafy vegetables in increasing serum concentrations of retinol and beta-carotene in schoolchildren in Indonesia", Am.J.Clin.Nutr ., vol. 68, no. 5, pp. 1058-1067. Deming, D. M., Boileau, A. C., Lee, C. M., & Erdman, J. W., Jr. 2000, "Amount of dietary fat and type of soluble fiber independently modulate postabsorptive conversion of beta-carotene to vitamin A in mongolian gerbils", J.Nutr ., vol. 130, no. 11, pp. 2789- 2796. Diaz-Bonilla, E. & Robinson, S. 2001, "Governments and public policy," in Shaping Globalization for Poverty Alleviation and Food Security, Focus 8 edn, E. Diaz-Bonilla & S. Robinson, eds., IFPRI, Washington D.C. Dijkhuizen, M. A., Wieringa, F. T., West, C. E., & Muhilal 2004, "Zinc plus beta- carotene supplementation of pregnant women is superior to beta-carotene supplementation alone in improving vitamin A status in both mothers and infants", Am.J.Clin.Nutr ., vol. 80, no. 5, pp. 1299-1307. Dimitrov, N. V ., Meyer, C., Ullrey, D. E., Chenoweth, W., Michelakis, A., Malone, W., Boone, C., & Fink, G. 1988, "Bioavailability of beta-carotene in humans", Am.J.Clin.Nutr ., vol. 48, no. 2, pp. 298-304. 57 Djunaedi, E., Sommer, A., Pandji, A., Kusdiono, & Taylor, H. R. 1988, "Impact of vitamin A supplementation on xerophthalmia. A randomized controlled community trial", Arch.Ophthalmol., vol. 106, no. 2, pp. 218-222. Dosti, M. P ., Mills, J. P ., Simon, P . W., & Tanumihardjo, S. A. 2006, "Bioavailability of beta-carotene (betaC) from purple carrots is the same as typical orange carrots while high-betaC carrots increase betaC stores in Mongolian gerbils (Meriones unguiculatus)", Br .J.Nutr ., vol. 96, no. 2, pp. 258-267. Edwards, A. J., Nguyen, C. H., You, C. S., Swanson, J. E., Emenhiser, C., & Parker, R. S. 2002, "Alpha- and beta-carotene from a commercial puree are more bioavailable to humans than from boiled-mashed carrots, as determined using an extrinsic stable isotope reference method", J.Nutr ., vol. 132, no. 2, pp. 159-167. Faber, M., Phungula, M. A., Venter, S. L., Dhansay, M. A., & Benade, A. J. 2002, "Home gardens focusing on the production of yellow and dark-green leafy vegetables increase the serum retinol concentrations of 2-5-y-old children in South Africa", Am.J.Clin.Nutr ., vol. 76, no. 5, pp. 1048-1054. FAO 2003, "FAOSTAT" [Cited 2008 January]. Available from: http://faostat.fao.org/site/502/default.aspx. Ferruzzi, M. G., Lumpkin, J. L., Schwartz, S. J., & Failla, M. 2006, "Digestive Stability, micellarization, and uptake of beta-carotene isomers by Caco-2 human intestinal cells", J.Agric.Food Chem., vol. 54, no. 7, pp. 2780-2785. Gaziano, J. M., Johnson, E. J., Russell, R. M., Manson, J. E., Stampfer, M. J., Ridker, P . M., Frei, B., Hennekens, C. H., & Krinsky, N. I. 1995, "Discrimination in absorption or transport of beta-carotene isomers after oral supplementation with either all-trans- or 9- cis-beta-carotene", Am.J.Clin.Nutr ., vol. 61, no. 6, pp. 1248-1252. Graham, J. M., Haskell, M. J., Pandey, P ., Shrestha, R. K., Brown, K. H., & Allen, L. H. 2007, "Supplementation with iron and riboflavin enhances dark adaptation response to vitamin A-fortified rice in iron-deficient, pregnant, nightblind Nepali women", Am.J.Clin.Nutr ., vol. 85, no. 5, pp. 1375-1384. Gronowska-Senger, A. & Wolf, G. 1970, "Effect of dietary protein on the enzyme from rat and human intestine which converts beta-carotene to retinal", J.Nutr ., vol. 100, no. 3, pp. 300-308. Guitérrez, E. D. 2006, "Modelaje del impacto de un programma de biofortificación en la adecuación de la ingesta de zinc en preescolares y mujeres en México", unpublished. Gurdian, M., Kontorovsky, I., Alvarado, E. A., Ramirez, S. A., & Hernández, R. 2005, Sistema Integrado de Vigilancia de Intervenciones Nutricionales (SIVIN). Informe de Progreso Nicaragua, 2003-2005 Ministerio de Salud (MINSA), Managua. 58 Haskell, M. J., Jamil, K. M., Hassan, F., Peerson, J. M., Hossain, M. I., Fuchs, G. J., & Brown, K. H. 2004, "Daily consumption of Indian spinach (Basella alba) or sweet potatoes has a positive effect on total-body vitamin A stores in Bangladeshi men", Am.J.Clin.Nutr ., vol. 80, no. 3, pp. 705-714. Haskell, M. J., Pandey, P ., Graham, J. M., Peerson, J. M., Shrestha, R. K., & Brown, K. H. 2005, "Recovery from impaired dark adaptation in nightblind pregnant Nepali women who receive small daily doses of vitamin A as amaranth leaves, carrots, goat liver, vitamin A-fortified rice, or retinyl palmitate", Am.J.Clin.Nutr ., vol. 81, no. 2, pp. 461- 471. Howe, J. A. & Tanumihardjo, S. A. 2006, "Carotenoid-biofortified maize maintains adequate vitamin a status in Mongolian gerbils", J.Nutr ., vol. 136, no. 10, pp. 2562-2567. Huiming, Y ., Chaomin, W., & Meng, M. 2005, "Vitamin A for treating measles in children", Cochrane.Database.Syst.Rev. no. 4, p. CD001479. ICBF 2006, Encuesta nacional de la situcación nutricional en Colombia 2005, 1 edn, Instituto Colombiano de Bienestar Familiar, Bogotá. IFPRI 2003, "PRAF: The Family Allowance Program" [Cited 2008 February]. Available from: http://www.ifpri.org/themes/praf.htm. IOM 2001, Dietary reference intakes for Vitamn A, Vitamin K, Arsenic, Boron, Chromium, Copper , Iodine, Iron, Manganese, Molybdenum, Nickel, Silicon, Vanadium, and Zinc National Academy Press, Washington D.C. Jalal, F., Nesheim, M. C., Agus, Z., Sanjur, D., & Habicht, J. P . 1998, "Serum retinol concentrations in children are affected by food sources of beta-carotene, fat intake, and anthelmintic drug treatment", Am.J.Clin.Nutr ., vol. 68, no. 3, pp. 623-629. Jones, G., Steketee, R. W., Black, R. E., Bhutta, Z. A., & Morris, S. S. 2003, "How many child deaths can we prevent this year?", Lancet, vol. 362, no. 9377, pp. 65-71. Katz, J., West, K. P ., Jr., Khatry, S. K., Pradhan, E. K., LeClerq, S. C., Christian, P ., Wu, L. S., Adhikari, R. K., Shrestha, S. R., & Sommer, A. 2000, "Maternal low-dose vitamin A or beta-carotene supplementation has no effect on fetal loss and early infant mortality: a randomized cluster trial in Nepal", Am.J.Clin.Nutr ., vol. 71, no. 6, pp. 1570-1576. Katz, J., West, K. P ., Jr., Khatry, S. K., Thapa, M. D., LeClerq, S. C., Pradhan, E. K., Pokhrel, R. P ., & Sommer, A. 1995, "Impact of vitamin A supplementation on prevalence and incidence of xerophthalmia in Nepal", Invest Ophthalmol.Vis.Sci., vol. 36, no. 13, pp. 2577-2583. Khan, N. C., West, C. E., de, P . S., Bosch, D., Phuong, H. D., Hulshof, P . J., Khoi, H. H., Verhoef, H., & Hautvast, J. G. 2007, "The contribution of plant foods to the vitamin A supply of lactating women in Vietnam: a randomized controlled trial", Am.J.Clin.Nutr ., vol. 85, no. 4, pp. 1112-1120. 59 Kongsbak, K., Wahed, M. A., Friis, H., & Thilsted, S. H. 2006, "Acute-phase protein levels, diarrhoea, Trichuris trichiura and maternal education are predictors of serum retinol: a cross-sectional study of children in a Dhaka slum, Bangladesh", Br .J.Nutr ., vol. 96, no. 4, pp. 725-734. Krause, V . M., Delisle, H., & Solomons, N. W. 1998, "Fortified foods contribute one half of recommended vitamin A intake in poor urban Guatemalan toddlers", J.Nutr ., vol. 128, no. 5, pp. 860-864. Lee, C. M., Lederman, J. D., Hofmann, N. E., & Erdman, J. W. 1998, "The mongolian gerbil (Meriones unguiculatus) is an appropriate animal model for evaluation of the conversion of beta-carotene to vitamin A", J.Nutr ., vol. 128, no. 2, pp. 280-286. Liu, C. S., Glahn, R. P ., & Liu, R. H. 2004, "Assessment of carotenoid bioavailability of whole foods using a Caco-2 cell culture model coupled with an in vitro digestion", J.Agric.Food Chem., vol. 52, no. 13, pp. 4330-4337. Low, J. W., Arimond, M., Osman, N., Cunguara, B., Zano, F., & Tschirley, D. 2007, "A food-based approach introducing orange-fleshed sweet potatoes increased vitamin A intake and serum retinol concentrations in young children in rural Mozambique", J.Nutr ., vol. 137, no. 5, pp. 1320-1327. Lozada, A. L., Flores, M., Rodriguez, S., & Barquera, S. 2007, "[Dietary patterns in Mexican adolescent girls. A comparison of two methods. National Nutrition Survey, 1999]", Salud Publica Mex., vol. 49, no. 4, pp. 263-273. Lutter, C. K. & Rivera, J. A. 2003, "Nutritional status of infants and young children and characteristics of their diets", J.Nutr ., vol. 133, no. 9, pp. 2941S-2949S. Mariath, J. G., Lima, M. C., & Santos, L. M. 1989, "Vitamin A activity of buriti (Mauritia vinifera Mart) and its effectiveness in the treatment and prevention of xerophthalmia", Am.J.Clin.Nutr ., vol. 49, no. 5, pp. 849-853. Mason, J., Bailes, A., Beda-Andourou, M., Copeland, N., Curtis, T., Deitchler, M., Foster, L., Hensley, M., Horjus, P ., Johnson, C., Lloren, T., Mendez, A., Munoz, M., Rivers, J., & Vance, G. 2005, "Recent trends in malnutrition in developing regions: vitamin A deficiency, anemia, iodine deficiency, and child underweight", Food Nutr .Bull., vol. 26, no. 1, pp. 59-108. McKevith, B. 2004, "Nutritional aspects of cereals. Final report to the home grown cereal authority" British Nutrition Foundation, London [Cited 2008 March]. Available from: www.hgca.com/publications/documents/cereals.pdf. Meenakshi, J. V ., Johnson, N., Manyog, V . M., De Groote, H., Javelosa, J., Yanggen, D., Naher, F., Gonzalez, C., Garcia, J., & Meng, E. 2007, "How cost-effective is biofortification in combating micronutrient malnutrition? An ex-ante assessment" [Cited 2008 September]. Available from: http://www.harvestplus.org/pdfs/hpwp02.pdf. 60 Mora, J. O., Gueri, M., & Mora, O. L. 1998, "Vitamin A deficiency in Latin America and the Caribbean: an overview", Rev.Panam.Salud Publica, vol. 4, no. 3, pp. 178-186. Munoz, E. C., Rosado, J. L., Lopez, P ., Furr, H. C., & Allen, L. H. 2000, "Iron and zinc supplementation improves indicators of vitamin A status of Mexican preschoolers", Am.J.Clin.Nutr ., vol. 71, no. 3, pp. 789-794. Murray, C. J. L. 1996, "Rethinking DALYs," in The global burden of disease, C. J. L. Murray & A. D. Lopez, eds., Harvard University Press, Cambridge. Murray, C. J. L. & Lopez, A. D. 1996, The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to 2020. Harvard University Press, Cambridge. Ncube, T. N., Greiner, T., Malaba, L. C., & Gebre-Medhin, M. 2001, "Supplementing lactating women with pureed papaya and grated carrots improved vitamin A status in a placebo-controlled trial", J.Nutr ., vol. 131, no. 5, pp. 1497-1502. Nestel, P ., Bouis, H. E., Meenakshi, J. V ., & Pfeiffer, W. 2006, "Biofortification of staple food crops", J.Nutr ., vol. 136, no. 4, pp. 1064-1067. Nestel, P ., Melara, A., Rosado, J., & Mora, J. O. 1999, "Vitamin A deficiency and anemia among children 12-71 months old in Honduras", Rev.Panam.Salud Publica, vol. 6, no. 1, pp. 34-43. PAHO 2007a, "Immunization newsletter" [Cited 2008 February]. Available from: http://www.paho.org/english/ad/fch/im/EPI_Newsletter.htm. PAHO 2007b, Health in the Americas, 2007 Pan American Health Organization, Washingon D.C. [Cited 2008 February]. Available from: http://www.paho.org/hia/homeing.html. Phillips, M., Sanghvi, T., Suarez, R., McKigney, J., & Fiedler, J. 1996, "The costs and effectiveness of three vitamin A interventions in Guatemala", Soc.Sci.Med., vol. 42, no. 12, pp. 1661-1668. Pinstrup-Andersen, P . 2007, "Agricultural research and policy for better health and nutrition in developing countries: a food systems approach", Agric.Econ., vol. 37, no. s1, pp. 187-198. Rahmathullah, L., Tielsch, J. M., Thulasiraj, R. D., Katz, J., Coles, C., Devi, S., John, R., Prakash, K., Sadanand, A. V ., Edwin, N., & Kamaraj, C. 2003, "Impact of supplementing newborn infants with vitamin A on early infant mortality: community based randomised trial in southern India", BMJ, vol. 327, no. 7409, p. 254. 61 Rahmathullah, L., Underwood, B. A., Thulasiraj, R. D., Milton, R. C., Ramaswamy, K., Rahmathullah, R., & Babu, G. 1990, "Reduced mortality among children in southern India receiving a small weekly dose of vitamin A", N.Engl.J.Med., vol. 323, no. 14, pp. 929-935. Ramakrishnan, U. & Darnton-Hill, I. 2002, "Assessment and control of vitamin A deficiency disorders", J.Nutr ., vol. 132, no. 9 Suppl, pp. 2947S-2953S. Rastogi, T. & Mathers, C. 2002, "Global burden of vitamin A deficiency in the year 2000", unpublished. Ribaya-Mercado, J. D., Maramag, C. C., Tengco, L. W., Dolnikowski, G. G., Blumberg, J. B., & Solon, F. S. 2007, "Carotene-rich plant foods ingested with minimal dietary fat enhance the total-body vitamin A pool size in Filipino schoolchildren as assessed by stable-isotope-dilution methodology", Am.J.Clin.Nutr ., vol. 85, no. 4, pp. 1041-1049. Ribaya-Mercado, J. D., Solomons, N. W., Medrano, Y ., Bulux, J., Dolnikowski, G. G., Russell, R. M., & Wallace, C. B. 2004, "Use of the deuterated-retinol-dilution technique to monitor the vitamin A status of Nicaraguan schoolchildren 1 y after initiation of the Nicaraguan national program of sugar fortification with vitamin A", Am.J.Clin.Nutr ., vol. 80, no. 5, pp. 1291-1298. Ribaya-Mercado, J. D., Solon, F. S., Solon, M. A., Cabal-Barza, M. A., Perfecto, C. S., Tang, G., Solon, J. A., Fjeld, C. R., & Russell, R. M. 2000, "Bioconversion of plant carotenoids to vitamin A in Filipino school-aged children varies inversely with vitamin A status", Am.J.Clin.Nutr ., vol. 72, no. 2, pp. 455-465. Rice, A. L., Stoltzfus, R. J., de, F. A., Chakraborty, J., Kjolhede, C. L., & Wahed, M. A. 1999, "Maternal vitamin A or beta-carotene supplementation in lactating bangladeshi women benefits mothers and infants but does not prevent subclinical deficiency", J.Nutr ., vol. 129, no. 2, pp. 356-365. Richelle, M., Enslen, M., Hager, C., Groux, M., Tavazzi, I., Godin, J. P ., Berger, A., Metairon, S., Quaile, S., Piguet-Welsch, C., Sagalowicz, L., Green, H., & Fay, L. B. 2004, "Both free and esterified plant sterols reduce cholesterol absorption and the bioavailability of beta-carotene and alpha-tocopherol in normocholesterolemic humans", Am.J.Clin.Nutr ., vol. 80, no. 1, pp. 171-177. Riedl, J., Linseisen, J., Hoffmann, J., & Wolfram, G. 1999, "Some dietary fibers reduce the absorption of carotenoids in women", J.Nutr ., vol. 129, no. 12, pp. 2170-2176. Rivera, J. A. & Sepulveda, A. J. 2003, "Conclusions from the Mexican National Nutrition Survey 1999: translating results into nutrition policy", Salud Publica Mex., vol. 45 Suppl 4, p. S565-S575. 62 Robles-Sardin, A. E., stiazaran-Garcia, H., valos-Navarro, R., Quihui-Cota, L., Cabrera- Pacheco, R. M., & Valencia, M. E. 1998, "[Effect of supplementation with a massive dose of vitamin A in children 6 to 36 months of age]", Salud Publica Mex., vol. 40, no. 4, pp. 309-315. Rock, C. L., Lovalvo, J. L., Emenhiser, C., Ruffin, M. T., Flatt, S. W., & Schwartz, S. J. 1998, "Bioavailability of beta-carotene is lower in raw than in processed carrots and spinach in women", J.Nutr ., vol. 128, no. 5, pp. 913-916. Rodriguez-Amaya, D. B. 1999a, "Changes in carotenoids during processing and storage of foods", Arch.Latinoam.Nutr ., vol. 49, no. 3 Suppl 1, pp. 38S-47S. Rodriguez-Amaya, D. B. 1999b, "Latin American food sources of carotenoids", Arch.Latinoam.Nutr ., vol. 49, no. 3 Suppl 1, pp. 74S-84S. Roodenburg, A. J., Leenen, R., van het Hof, K. H., Weststrate, J. A., & Tijburg, L. B. 2000, "Amount of fat in the diet affects bioavailability of lutein esters but not of alpha- carotene, beta-carotene, and vitamin E in humans", Am.J.Clin.Nutr ., vol. 71, no. 5, pp. 1187-1193. Ross, A. C. 2006, "Vitamin A and carotenoids," in Modern nutrition in health and disease, 10 edn, M. E. Shills et al., eds., Lippincott Williams & Wilkins, Baltimore, pp. 351-375. Ryan, L., O'Connell, O., O'Sullivan, L., Aherne, S. A., & O'Brien, N. M. 2008, "Micellarisation of carotenoids from raw and cooked vegetables", Plant Foods Hum.Nutr ., vol. 63, no. 3, pp. 127-133. Schultink, W. 2002, "Use of under-five mortality rate as an indicator for vitamin A deficiency in a population", J.Nutr ., vol. 132, no. 9 Suppl, pp. 2881S-2883S. SCN 2004, 5th report on the world nutrition situation United Nations, Geneva [Cited 2008 February]. Available from: http://www.unsystem.org/scn/publications/AnnualMeeting/SCN31/SCN5Report.pdf. Shankar, A. V ., West, K. P ., Jr., Gittelsohn, J., Katz, J., & Pradhan, R. 1996, "Chronic low intakes of vitamin A-rich foods in households with xerophthalmic children: a case- control study in Nepal", Am.J.Clin.Nutr ., vol. 64, no. 2, pp. 242-248. Sinha, D. P . & Bang, F. B. 1976, "The effect of massive doses of vitamin A on the signs of vitamin A deficiency in preschool children", Am.J.Clin.Nutr ., vol. 29, no. 1, pp. 110- 115. Sommer, A. & Davidson, F. R. 2002, "Assessment and control of vitamin A deficiency: the Annecy Accords", J.Nutr ., vol. 132, no. 9 Suppl, pp. 2845S-2850S. 63 Sommer, A., Katz, J., & Tarwotjo, I. 1984, "Increased risk of respiratory disease and diarrhea in children with preexisting mild vitamin A deficiency", Am.J.Clin.Nutr ., vol. 40, no. 5, pp. 1090-1095. Sommer, A. & Tarwotjo, I. 1982, "Protein deficiency and treatment of xerophthalmia", Arch.Ophthalmol., vol. 100, no. 5, pp. 785-787. Sommer, A., Tarwotjo, I., Djunaedi, E., West, K. P ., Jr., Loeden, A. A., Tilden, R., & Mele, L. 1986, "Impact of vitamin A supplementation on childhood mortality. A randomised controlled community trial", Lancet, vol. 1, no. 8491, pp. 1169-1173. Sommer, A., Tarwotjo, I., Hussaini, G., & Susanto, D. 1983, "Increased mortality in children with mild vitamin A deficiency", Lancet, vol. 2, no. 8350, pp. 585-588. Sommer, A., Tarwotjo, I., & Katz, J. 1987, "Increased risk of xerophthalmia following diarrhea and respiratory disease", Am.J.Clin.Nutr ., vol. 45, no. 5, pp. 977-980. Sommer, A. & West, K. P . 1996,V itamin A deficiency: health survival and vision Oxford University Press, New York. Sovani, I., Humphrey, J. H., Kuntinalibronto, D. R., Natadisastra, G., Muhilal, & Tielsch, J. M. 1994, "Response of Bitot's spots to a single oral 100,000- or 200,000-IU dose of vitamin A", Am.J.Ophthalmol., vol. 118, no. 6, pp. 792-796. Stein, A. J., Meenakshi, J. V ., Qaim, M., Nestel, P ., Sachdev, H. P . S., & Bhutta, Z. A. 2005, Analyzing the health benefits of biofortified staple crops by means of the Disability- Adjusted Life Years approach: a handbook focusing on iron, zinc and vitamin A HarvestPlus, Washington. Stein, A. J., Sachdev, H. P . S., & Qaim, M. 2006, "Potential impact and cost- effectiveness of Golden Rice", Nat.Biotechnol., vol. 24, no. 10, pp. 1200-1201. Stoltzfus, R. J., Hakimi, M., Miller, K. W., Rasmussen, K. M., Dawiesah, S., Habicht, J. P ., & Dibley, M. J. 1993, "High dose vitamin A supplementation of breast-feeding Indonesian mothers: effects on the vitamin A status of mother and infant", J.Nutr ., vol. 123, no. 4, pp. 666-675. Sulaeman, A., Giraud, D. W., Naslund, M. M., & Driskell, J. A. 2002, "Mongolian gerbils can utilize provitamin-A carotenoids in deep-fried carrot chips", J.Nutr ., vol. 132, no. 2, pp. 211-217. Tang, G., Gu, X., Hu, S., Xu, Q., Qin, J., Dolnikowski, G. G., Fjeld, C. R., Gao, X., Russell, R. M., & Yin, S. 1999, "Green and yellow vegetables can maintain body stores of vitamin A in Chinese children", Am.J.Clin.Nutr ., vol. 70, no. 6, pp. 1069-1076. Tang, G., Qin, J., Dolnikowski, G. G., Russell, R. M., & Grusak, M. A. 2005, "Spinach or carrots can supply significant amounts of vitamin A as assessed by feeding with intrinsically deuterated vegetables", Am.J.Clin.Nutr ., vol. 82, no. 4, pp. 821-828. 64 Tanumihardjo, S. A. 2004, "Assessing vitamin A status: past, present and future", J.Nutr ., vol. 134, no. 1, pp. 290S-293S. Tanumihardjo, S. A. & Howe, J. A. 2005, "Twice the amount of alpha-carotene isolated from carrots is as effective as beta-carotene in maintaining the vitamin A status of Mongolian gerbils", J.Nutr ., vol. 135, no. 11, pp. 2622-2626. Tarwotjo, I., Sommer, A., Soegiharto, T., Susanto, D., & Muhilal 1982, "Dietary practices and xerophthalmia among Indonesian children", Am.J.Clin.Nutr ., vol. 35, no. 3, pp. 574-581. Thatcher, A. J., Lee, C. M., & Erdman, J. W., Jr. 1998, "Tissue stores of beta-carotene are not conserved for later use as a source of vitamin A during compromised vitamin A status in mongolian gerbils (Meriones unguiculatus)", J.Nutr ., vol. 128, no. 7, pp. 1179-1185. Tielsch, J. M., Rahmathullah, L., Katz, J., Thulasiraj, R. D., Coles, C., Sheeladevi, S., & Prakash, K. 2008, "Maternal night blindness during pregnancy is associated with low birthweight, morbidity, and poor growth in South India", J.Nutr ., vol. 138, no. 4, pp. 787- 792. Torney, F., Moeller, L., Scarpa, A., & Wang, K. 2007, "Genetic engineering approaches to improve bioethanol production from maize", Curr .Opin.Biotechnol., vol. 18, pp. 193- 199. Underwood, B. A. 2004, "Vitamin A deficiency disorders: international efforts to control a preventable "pox"", J.Nutr ., vol. 134, no. 1, pp. 231S-236S. UNDP 2008, "Global human development report country information" [Cited 2008 March]. Available from: http://hdr.undp.org/en/countries/. UNICEF 1997, State of the world's children 1998 United Nations Children's Fund, New York [Cited 2008 August]. Available from:h ttp://www.unicef.org/sowc98/. UNICEF 2007, State of the world's children 2008: child survival United Nations Children's Fund, New York [Cited 2008 February]. Available from: http://www.unicef.org/sowc08/report/report.php. van Jaarsveld, P . J., Faber, M., Tanumihardjo, S. A., Nestel, P ., Lombard, C. J., & Benade, A. J. 2005, "Beta-carotene-rich orange-fleshed sweet potato improves the vitamin A status of primary school children assessed with the modified-relative-dose- response test", Am.J.Clin.Nutr ., vol. 81, no. 5, pp. 1080-1087. van Vliet, T., van Vlissingen, M. F., van, S. F., & van den, B. H. 1996, "beta-Carotene absorption and cleavage in rats is affected by the vitamin A concentration of the diet", J.Nutr ., vol. 126, no. 2, pp. 499-508. 65 Vijayaraghavan, K., Sarma, K. V ., Rao, N. P ., & Reddy, V . 1984, "Impact of massive doses of vitamin A on incidence of nutritional blindness", Lancet, vol. 2, no. 8395, pp. 149-151. Villard, L. & Bates, C. J. 1986, "Carotene dioxygenase (EC 1.13.11.21) activity in rat intestine: effects of vitamin A deficiency and of pregnancy", Br .J.Nutr ., vol. 56, no. 1, pp. 115-122. West, K. P . 2002, "Extent of vitamin A deficiency among preschool children and women of reproductive age", J.Nutr ., vol. 132, no. 9 Suppl, pp. 2857S-2866S. West, K. P ., Katz, J., Khatry, S. K., LeClerq, S. C., Pradhan, E. K., Shrestha, S. R., Connor, P . B., Dali, S. M., Christian, P ., Pokhrel, R. P ., & Sommer, A. 1999, "Double blind, cluster randomised trial of low dose supplementation with vitamin A or beta carotene on mortality related to pregnancy in Nepal. The NNIPS-2 Study Group", BMJ, vol. 318, no. 7183, pp. 570-575. West, K. P ., Pokhrel, R. P ., Katz, J., LeClerq, S. C., Khatry, S. K., Shrestha, S. R., Pradhan, E. K., Tielsch, J. M., Pandey, M. R., & Sommer, A. 1991, "Efficacy of vitamin A in reducing preschool child mortality in Nepal", Lancet, vol. 338, no. 8759, pp. 67-71. WHO 1996, Indicators for assessing vitamin A deficiency and their application in monitoring and evaluating intervention programmes World Health Organization, Geneva [Cited 2008 January]. Available from: http://www.who.int/nutrition/topics/vad/en/. WHO 2001, Macroeconomics and health: investing in health for economic development World Health Organization, Geneva [Cited 2008 June]. Available from: http://www.who.int.cmhreport/. WHO 2006, Neonatal and perinatal mortality: country, regional and global estimates World Health Organization, Geneva [Cited 2008 March]. Available from: http://www.who.int/making_pregnancy_safer/publications/neonatal.pdf. WHO 2008a, "Vitamin and Mineral Nutrition Information System" [Cited 2008 January]. Available from: http://www.who.int/vmnis/vitamina/data/database/countries/en/index.html. WHO 2008b, "Global Data Bank on Infant and Young Child Feeding" [Cited 2008 February]. Available from: http://www.who.int/nutrition/databases/infantfeeding/en/index.html. Wieringa, F. T., Dijkhuizen, M. A., West, C. E., Thurnham, D. I., Muhilal, & Van der Meer, J. W. 2003, "Redistribution of vitamin A after iron supplementation in Indonesian infants", Am.J.Clin.Nutr ., vol. 77, no. 3, pp. 651-657. 66 World Bank 2001, "Living Standard Measurement Survey" [Cited 2008 January]. Available from: http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/EXTLSM S/0,,contentMDK:21588637~isCURL:Y~menuPK:4196952~pagePK:64168445~piPK:64 168309~theSitePK:3358997,00.html. Yeum, K. J. & Russell, R. M. 2002, "Carotenoid bioavailability and bioconversion", Annu.Rev.Nutr ., vol. 22, pp. 483-504. Zimmermann, R. & Qaim, M. 2004, "Potential health benefits of Golden Rice: a Philippine case study", Food Policy, vol. 29, pp. 147-168. 67 Appendix A: Applied prevalence of VAD outcomes Estimates of prevalence of XN in women of reproductive age and estimates of XS, X3A, X3B, XS, measles, blindness and mortality in children < 5 years of age in the respective countries. Women Children (a) Xerophthalmia Measles (b) Blindness Mortality Country XN XN XS No. confirmed (%) (%) per 1000 birthsΨ Colombia 3.7 0.1275† 0.0026 0 0 0.001275 21 Honduras 4.8β 0.3137§ 0.0063 0 0 0.003137 27 Mexico 0£ 0.0000£ 0.0000 23 0.000002202 0.000000 35 Nicaragua 1.9 0.3137§ 0.0063 0 0 0.003137 36 (a) Data are from West (2002) (b) Data are from PAHO (2007a) β Data from the 2005-2006 national survey (WHO 2008b) £ Data obtained from consulting J. R. Dommarco † Data extrapolated from national weighted estimates for Brazil (West 2002). § Data extrapolated from national weighted estimates for Egypt (West 2002) Ψ Data were derived from UNICEF (2007) 68 Appendix B: Applied data on VA intake 69 Overview of the nutrition surveys applied either directly or for comparison purposes in the lack of primary data for children <5 years of age and women of reproductive age. Age Median dietary Mean cassava Mean maize Country Year Survey Samplesize group Sex Dietary (n) Method Vitamin A intake c o n s ump-tion consump-tion Source (years) (μg RAE/d) (g/day) (g/day) Colombia 2005 N 10,913 2-3 B 24-H 707.3 ICBF 2006 2005 N NA 19-50 F 24-H NA ICBF 2006 2005 N 3,560 14-18 F 24-H 636.0* ICBF 2006 2005 N 1,084 51-64 F 24-H 656.0* ICBF 2006 2005 N 4,501 2-4 B 24-H 10.8 17.3 ICBF 2006 2005 N 8,544 15-49 F 24-H 18.2 31.5 ICBF 2006 Honduras 1999-2003 N 6,648 <5 B HH-FFQ 152.1* 26.4 162.9 IFPRI 2003 1999-2003 N 8,214 >14 F HH-FFQ 147.8* 26.4 172.5 IFPRI 2003 Mexico 1999 N 2,630 12-49 F 24-H 360.3* NA Barquera et al. 2003a 1999 N 1,072 1-4 B 24-H 265.1* NA Barquera et al. 2003b 2003 N FBS 0.6 FAO 2003 1999 N 2,630 12-49 F 24-H 27.0 Guitérrez 2006 1999 N 1,072 1-4 B 24-H 85.2 Guitérrez 2006 Nicaragua 2001 N 3,435 <5 B HH-FFQ 439.0* 133.1 91.3 World Bank 2001 2001 N 6,608 >14 F HH-FFQ 587.9* 59.6 113.0 World Bank 2001 Survey: N=national, S=subsample. Sex: B=both, F=female. Dietary Method: 24-H=24-hour dietary recall, HH-FFQ=household food frequency questionnaire, FBS=Food Balance Sheets *Does not meet dietary recommendations V A according to the RDA values applied from United States and Canada (IOM 2001) The food composition table used by HarvestPlus for calculating the V A consumed in Honduras and Nicaragua was obtained from FAO and presented in mean values. Appendix C: Data sources for the DALYs calculations Colombia Data Source Population size, CBR, U5MR, GNP, SLE (UNICEF 2007) Percentage of children ever breastfed (ICBF 2006) Perinatal deaths, live births and still births (WHO 2006) Childhood X (West 2002) Brazil Maternal X (West 2002) Measles incidence (personal communication, E. Poveda) Mean daily intake of vitamin A, children (ICBF 2006) age 2-3 years Mean daily intake of vitamin A, women (ICBF 2006) - estimated mean consumption of V A applying data from age 14-18 and 51-64 Mean daily intake of cassava Data derived from ENSIN1 2005 Mean daily intake of maize Data derived from ENSIN 2005 1 Encuesta nacional de la situación nutricional en Colombia (ENSIN) is Colombia’s national nutrition collection carried out in 2005. Access was given to statistically unprocessed data from this survey, which can not be found in the publication made by Instituto Colombiano de Bienestar Familiar (ICBF 2006) 70 Honduras Data Source Population size, CBR, U5MR, GNP, SLE (UNICEF 2007) Percentage of children ever breastfed (WHO 2008a) 1996 Perinatal deaths, live births and still births (WHO 2006) Childhood X (West 2002) Egypt Maternal X (WHO 2008b) 2005-2006 Measles incidence (PAHO 2007a) Mean daily intake of vitamin A, children (HarvestPlus)2 IFPRI 2003 Mean daily intake of vitamin A, women (HarvestPlus) IFPRI 2003 Mean daily intake of cassava (HarvestPlus) IFPRI 2003 Mean daily intake of maize (HarvestPlus) IFPRI 2003 2 With respect to the data processed by HarvestPlus, these were obtained from James García, CIAT. 71 Mexico Data Source Population size, CBR, U5MR, GNP, SLE (UNICEF 2007) Percentage of children ever breastfed (WHO 2008a) DHS 1988 Perinatal deaths, live births and still births (WHO 2006) Childhood X (personal communication, J. R. Dommarco) Maternal X (personal communication, J. R. Dommarco) Measles incidence (PAHO 2007a) Mean daily intake of vitamin A, children (Barquera et al. 2003b) Mean daily intake of vitamin A, women (Barquera et al. 2003a) Mean daily intake of cassava (FAO 2003) Mean daily intake of maize (Guitérrez 2006) 72 Nicaragua Data Source Population size, CBR, U5MR, GNP, SLE (UNICEF 2007) Percentage of children ever breastfed (WHO 2008a) DHS 2001 Perinatal deaths, live births and still births (WHO 2006) Childhood X (West 2002) Egypt Maternal X (West 2002) Measles incidence (PAHO 2007a) Mean daily intake of vitamin A, children (HarvestPlus)3 World Bank 2001 Mean daily intake of vitamin A, women (HarvestPlus) World Bank 2001 Mean daily intake of cassava (HarvestPlus) World Bank 2001 Mean daily intake of maize (HarvestPlus) World Bank 2001 3 With respect to the data processed by HarvestPlus, these were obtained from James García, CIAT. 73 Appendix D: Assumptions related to the two scenarios Cassava Data Source Beta-carotene content, optimistic (personal communication, S.S. Pérez) Beta-carotene content, pessimistic (personal communication, S.S. Pérez) Bioconversion, optimistic (Stein et al. 2005) Bioconversion, pessimistic (Stein et al. 2005) Post harvest losses, optimistic (Chávez et al. 2008) Post harvest losses, pessimistic (Chávez et al. 2008) Technology coverage rate, optimistic (Meenakshi et al. 2007) Northeast Brazil Technology coverage rate, pessimistic (Meenakshi et al. 2007) Northeast Brazil Assumptions related to the DALYs calculations. Cassava Parameter of assumption Optimistic Pessimistic β-carotene content (µg/gr) 14.5 7.5 Bioconversion rate to V A (x:1) 12 12 Post-harvest losses (%) 28.1 82.7 Technology coverage rate (%) 40 15 β-carotene content refers to the difference in current content and the target contents. Current β-carotene content of cassava is set to 0.5 µg/g wet weight. 74 Maize Data Source Beta-carotene content, optimistic (personal communication, H. Pachón) Beta-carotene content, pessimistic (personal communication, H. Pachón) Bioconversion, optimistic (Stein et al. 2005) Bioconversion, pessimistic (Stein et al. 2005) Post harvest losses, optimistic (Zimmermann & Qaim 2004) Post harvest losses, pessimistic (McKevith 2004) Technology coverage rate, optimistic (Zimmermann & Qaim 2004) Technology coverage rate, pessimistic (Zimmermann & Qaim 2004) Assumptions related to the DALYs calculations. Maize Parameter of assumption Optimistic Pessimistic β-carotene content (µg/gr) 15.0 10.0 Bioconversion rate to V A (x:1) 12 12 Post-harvest losses (%) 0.0 66.7 Technology coverage rate (%) 50 25 β-carotene content refers to the difference in current content and the target contents. Current β-carotene content of maize is set to 0.5 µg/g dry weight. 75 Appendix E: Efficacy tables Efficacy of introducing pro-vitamin A biofortified cassava in Colombia with conditions of the optimistic scenario Children Pregnant Lactating <5 years Women Women Cassava intake (g/day) 12.7 17.3 17.3 Current V A supply from all food sources (µg/day) CV A 707.3 646.0 646.0 RDA4 for VA (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 0.0 154.0 604.0 Beta-carotene intake through cassava (µg/day) 132.1 180.5 180.5 V A from biofortified cassava after bioconversion (µg) 11.0 15.0 15.0 Improved VA supply (µg/day) IV A 718.3 661.0 661.0 Contribution of cassava to reduce VA deficit (%) 0.00 9.77 2.49 Efficacy (%) 0.00 19.78 6.21 Efficacy of introducing pro-vitamin A biofortified cassava in Colombia with conditions of the pessimistic scenario Children Pregnant Lactating <5 years Women Women Cassava intake (g/day) 12.7 17.3 17.3 Current V A supply from all food sources (µg/day) CV A 707.3 646.0 646.0 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 0.0 154.0 604.0 Beta-carotene intake through cassava (µg/day) 16.4 22.4 22.4 V A from cassava after bioconversion (µg) 1.4 1.9 1.9 Improved VA supply (µg/day) IV A 708.7 647.9 647.9 Contribution of cassava to reduce VA deficit (%) 0.00 1.21 0.31 Efficacy (%) 0.00 2.60 0.79 4 RDA is the point of average nutrient intake where the greatest majority (97.5%) of the population is not at risk of developing a deficiency. The RDAs applied in these analyses correspond to those applied by Zimmermann & Qaim (2004), however, their values not completely in accordance with those reported by the IOM (2001). 76 Efficacy of introducing pro-vitamin A biofortified cassava in Honduras with conditions of the optimistic scenario Children Pregnant Lactating <5 years Women Women Cassava intake (g/day) 26.4 26.4 26.4 Current V A supply from all food sources (µg/day) CV A 152.1 147.8 147.8 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 347.9 652.2 1,102.2 Beta-carotene intake through cassava (µg/day) 275.2 275.2 275.2 V A from biofortified cassava after bioconversion (µg) 22.9 22.9 22.9 Improved VA supply (µg/day) IV A 175.0 170.7 170.7 Contribution of cassava to reduce VA deficit (%) 6.59 3.52 2.08 Efficacy (%) 19.13 13.23 10.05 Efficacy of introducing pro-vitamin A biofortified cassava in Honduras with conditions of the pessimistic scenario Children Pregnant Lactating <5 years Women Women Cassava intake (g/day) 26.4 26.4 26.4 Current V A supply from all food sources (µg/day) CV A 152.1 147.8 147.8 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 347.9 652.2 1,102.2 Beta-carotene intake through cassava (µg/day) 34.2 34.2 34.2 V A from biofortified cassava after bioconversion (µg) 2.9 2.9 2.9 Improved VA supply (µg/day) IV A 155.0 150.7 150.7 Contribution of cassava to reduce VA deficit (%) 0.82 0.44 0.26 Efficacy (%) 2.60 1.78 1.34 Efficacy of introducing pro-vitamin A biofortified cassava in Mexico with conditions of the optimistic scenario Children Pregnant Lactating <5 years Women Women Cassava intake (g/day) 0.6 0.6 0.6 Current V A supply from all food sources (µg/day) CV A 265.1 360.3 360.3 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 234.9 439.7 889.7 Beta-carotene intake through cassava (µg/day) 6.0 6.0 6.0 V A from biofortified cassava after bioconversion (µg) 0.5 0.5 0.5 Improved VA supply (µg/day) IV A 265.6 360.8 360.8 Contribution of cassava to reduce VA deficit (%) 0.21 0.11 0.06 Efficacy (%) 0.54 0.31 0.19 77 Efficacy of introducing pro-vitamin A biofortified cassava in Mexico with conditions of the pessimistic scenario Children Pregnant Lactating <5 years Women Women Cassava intake (g/day) 0.6 0.6 0.6 Current V A supply from all food sources (µg/day) CV A 265.1 360.3 360.3 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 234.9 439.7 889.7 Beta-carotene intake through cassava (µg/day) 0.8 0.8 0.8 V A from biofortified cassava after bioconversion (µg) 0.1 0.1 0.1 Improved VA supply (µg/day) IV A 265.2 360.4 360.4 Contribution of cassava to reduce VA deficit (%) 0.03 0.01 0.01 Efficacy (%) 0.07 0.04 0.02 Efficacy of introducing pro-vitamin A biofortified cassava in Nicaragua with conditions of the optimistic scenario Children Pregnant Lactating <5 years Women Women Cassava intake (g/day) 113.0 59.6 59.6 Current V A supply from all food sources (µg/day) CV A 439.0 587.9 587.9 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 61.0 212.1 662.1 Beta-carotene intake through cassava (µg/day) 1,177.9 621.3 621.3 V A from biofortified cassava after bioconversion (µg) 98.2 51.8 51.8 Improved VA supply (µg/day) IV A 537.2 639.7 639.7 Contribution of cassava to reduce VA deficit (%) 160.92 24.41 7.82 Efficacy (%) 100.00 45.85 19.13 Efficacy of introducing pro-vitamin A biofortified cassava in Nicaragua with conditions of the pessimistic scenario Children Pregnant Lactating <5 years Women Women Cassava intake (g/day) 113.0 59.6 59.6 Current V A supply from all food sources (µg/day) CV A 439.0 587.9 587.9 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 61.0 212.1 662.1 Beta-carotene intake through cassava (µg/day) 146.4 77.2 77.2 V A from biofortified cassava after bioconversion (µg) 12.2 6.4 6.4 Improved VA supply (µg/day) IV A 451.2 594.3 594.3 Contribution of cassava to reduce VA deficit (%) 20.01 3.03 0.97 Efficacy (%) 37.15 6.62 2.55 78 Efficacy of introducing pro-vitamin A biofortified maize in Colombia with conditions of the optimistic scenario Children Pregnant Lactating <5 years Women Women Maize intake (g/day) 12.7 17.4 17.4 Current V A supply from all food sources (µg/day) CV A 707.3 646.0 646.0 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 0.0 154.0 604.0 Beta-carotene intake through maize (µg/day) 190.1 260.6 260.6 V A from maize after bioconversion (µg) 15.8 21.7 21.7 Improved VA supply (µg/day) IV A 723.1 667.7 667.7 Contribution of maize to reduce VA deficit (%) 0.0 14.1 3.59 Efficacy (%) 0.0 27.8 8.87 Efficacy of introducing pro-vitamin A biofortified maize in Colombia with conditions of the pessimistic scenario Children Pregnant Lactating <5 years Women Women Maize intake (g/day) 12.7 17.4 17.4 Current V A supply from all food sources (µg/day) CV A 707.3 646.0 646.0 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 0.0 154.0 604.0 Beta-carotene intake through maize (µg/day) 42.2 57.9 57.9 V A from maize after bioconversion (µg) 3.5 4.8 4.8 Improved VA supply (µg/day) IV A 710.8 650.8 650.8 Contribution of maize to reduce VA deficit (%) 0.00 3.13 0.80 Efficacy (%) 0.00 6.62 2.02 Efficacy of introducing pro-vitamin A biofortified maize in Honduras with conditions of the optimistic scenario Children Pregnant Lactating <5 years Women Women Maize intake (g/day) 162.9 172.5 172.5 Current V A supply from all food sources (µg/day) CV A 152.1 147.8 147.8 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 347.9 652.2 1,102.2 Beta-carotene intake through maize (µg/day) 2,443.5 2,587.5 2,587.5 V A from maize after bioconversion (µg) 203.6 215.6 215.6 Improved VA supply (µg/day) IV A 355.7 363.4 363.4 Contribution of maize to reduce VA deficit (%) 58.53 33.06 19.56 Efficacy (%) 89.50 72.14 58.02 79 Efficacy of introducing pro-vitamin A biofortified maize in Honduras with conditions of the pessimistic scenario Children Pregnant Lactating <5 years Women Women Maize intake (g/day) 162.9 172.5 172.5 Current V A supply from all food sources (µg/day) CV A 152.1 147.8 147.8 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 347.9 652.2 1,102.2 Beta-carotene intake through maize (µg/day) 543.1 575.1 575.1 V A from maize after bioconversion (µg) 45.3 47.9 47.9 Improved VA supply (µg/day) IV A 197.4 195.7 195.7 Contribution of maize to reduce VA deficit (%) 13.01 7.35 4.35 Efficacy (%) 34.39 25.59 9.35 Efficacy of introducing pro-vitamin A biofortified maize in Mexico with conditions of the optimistic scenario Children Pregnant Lactating <5 years Women Women Maize intake (g/day) 27.0 85.2 85.2 Current V A supply from all food sources (µg/day) CV A 265.1 360.3 360.3 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 234.9 439.7 889.7 Beta-carotene intake through maize (µg/day) 405.0 1,278.0 1,278.0 V A from maize after bioconversion (µg) 33.8 106.5 106.5 Improved VA supply (µg/day) IV A 298.9 466.8 466.8 Contribution of maize to reduce VA deficit (%) 14.37 24.22 11.97 Efficacy (%) 31.78 50.73 32.65 Efficacy of introducing pro-vitamin A biofortified maize in Mexico with conditions of the pessimistic scenario Children Pregnant Lactating <5 years Women Women Maize intake (g/day) 27.0 85.2 85.2 Current V A supply from all food sources (µg/day) CV A 265.1 360.3 360.3 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 234.9 439.7 889.7 Beta-carotene intake through maize (µg/day) 90.0 284.1 284.1 V A from maize after bioconversion (µg) 7.5 23.7 23.7 Improved VA supply (µg/day) IV A 272.6 384.0 384.0 Contribution of maize to reduce VA deficit (%) 3.19 5.38 2.66 Efficacy (%) 7.83 13.72 8.40 80 Efficacy of introducing pro-vitamin A biofortified maize in Nicaragua with conditions of the optimistic scenario Children Pregnant Lactating <5 years Women Women Maize intake (g/day) 91.3 113.0 113.0 Current V A supply from all food sources (µg/day) CV A 439.0 587.9 587.9 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 61.0 212.1 662.1 Beta-carotene intake through maize (µg/day) 1,369.5 1,695.0 1,695.0 V A from maize after bioconversion (µg) 114.1 141.3 141.3 Improved VA supply (µg/day) IV A 553.1 729.2 729.2 Contribution of maize to reduce VA deficit (%) 187.09 66.60 21.33 Efficacy (%) 100.00 90.29 45.55 Efficacy of introducing pro-vitamin A biofortified maize in Nicaragua with conditions of the pessimistic scenario Children Pregnant Lactating <5 years Women Women Maize intake (g/day) 91.3 113.0 113.0 Current V A supply from all food sources (µg/day) CV A 439.0 587.9 587.9 RDA for V A (µg/day) RDA 500.0 800.0 1,250.0 V A deficit (µg/day) 61.0 212.1 662.1 Beta-carotene intake through maize (µg/day) 304.4 376.7 376.7 V A from maize after bioconversion (µg) 25.4 31.4 31.4 Improved VA supply (µg/day) IV A 464.4 619.3 619.3 Contribution of maize to reduce VA deficit (%) 41.58 14.80 4.74 Efficacy (%) 67.11 29.77 11.98 81 Appendix F: New incidence rates New incidence rates and number of people affected by V AD in the optimistic and pessimistic scenarios in Colombia after the introduction of pro-V A-biofortified cassava. Optimistic Pessimistic New Incidence New Number New Clinical manifestation of affected Incidence New Number rate of affected (%) (thousand) rate (%) (thousand) XN Children <=5 0.13% 5.7 0.13% 5.7 Pregnant women 3.41% 30.1 3.69% 32.6 Lactating women 3.61% 31.0 3.70% 31.7 New incidence rates and number of people affected by V AD in the optimistic and pessimistic scenarios in Honduras after the introduction of pro-V A-biofortified cassava. Optimistic Pessimistic New Incidence New Number New New Number Clinical manifestation of affected Incidence rate rate of affected (%) (thousand) (%) (thousand) XN Children <=5 0.29% 2.7 0.31% 2.9 Pregnant women 4.55% 9.0 4.79% 9.5 Lactating women 4.61% 8.8 4.80% 9.2 New incidence rates and number of people affected by V AD in the optimistic and pessimistic scenarios in Nicaragua after the introduction of pro-V A-biofortified cassava. Optimistic Pessimistic New New Number New New Number Clinical manifestation Incidence rate of affected Incidence rate of affected (%) (thousand) (%) (thousand) XN Children <=5 0.19% 1.3 0.30% 2.0 Pregnant women 1.55% 2.2 1.88% 2.6 Lactating women 1.75% 2.3 1.90% 2.5 82 New incidence rates and number of people affected by V AD in the optimistic and pessimistic scenarios in Colombia after the introduction of pro-V A-biofortified maize. Optimistic Pessimistic New New Clinical manifestation Incidence New Number of affected Incidence New Number rate (thousand) rate of affected (%) (%) (thousand) XN Children <=5 0.13% 5.7 0.13% 5.7 Pregnant women 3.19% 28.2 3.64% 32.2 Lactating women 3.54% 30.4 3.70% 31.6 New incidence rates and number of people affected by V AD in the optimistic and pessimistic scenarios in Honduras after the introduction of pro-V A-biofortified maize. Optimistic Pessimistic New New Number New New Number Clinical manifestation Incidence rate of affected Incidence rate of affected (%) (thousand) (%) (thousand) XN Children <=5 0.17% 1.6 0.29% 2.7 Pregnant women 3.07% 6.1 4.50% 8.9 Lactating women 3.41% 6.5 4.60% 8.7 New incidence rates and number of people affected by V AD in the optimistic and pessimistic scenarios in Nicaragua after the introduction of pro-V A-biofortified maize. Optimistic Pessimistic New New Clinical manifestation Incidence New Number New Number rate of affected Incidence of affected (%) (thousand) rate (%) (thousand) XN Children <=5 0.16% 1.1 0.26% 1.8 Pregnant women 1.04% 1.4 1.76% 2.4 Lactating women 1.47% 1.9 1.80% 2.4 83 Appendix G: Sensitivity results of the extrapolated <5 XN Calculated loss of DALYs per year due to V AD without introduction of the pro-V A-biofortified cassava or maize applying the extrapolated prevalence of childhood X in Colombia, Honduras and Nicaragua. Colombia Honduras Mexico Nicaragua Loss Total Total Total Total due to mortality 15,772 4,491 64,774 4,303 due to temporary disability 3,225 900 0 268 due to permanent disability 144 25 0 18 Total due to disability 3,369 925 0 286 Total 19,141 5,416 64,774 4,589 In monetary terms (million USD) 52.0 6.5 509.8 4.6 Calculated loss of DALYs per year due to V AD without introduction of the pro-V A-biofortified cassava or maize applying a zero prevalence of childhood X in Colombia, Honduras and Nicaragua. Loss Colombia Honduras Mexico Nicaragua Total Total Total Total due to mortality 15,772 4,491 64,774 4,303 due to temporary disability 2,941 854 0 235 due to permanent disability 0 0 0 0 Total due to disability 2,941 854 0 235 Total 18,713 5,345 64,774 4,538 In monetary terms (million USD) 51.0 6.4 509.8 4.5 Sensitivity to the extrapolated prevalence of X expressed as absolute and relative differences in the amount of DALYs and USD lost due to V AD each year in Colombia, Honduras and Nicaragua. Colombia Honduras Mexico Nicaragua Absolute ΔDALYs lost 428 71 0 51 % decrease in total DALYs lost 2.2 1.3 0.0 1.1 Absolute Δmonetary loss 1.1 0.1 0.0 0.1 % decrease in monetary loss 2.1 1.5 0.0 2.2 84 Appendix H: Sensitivity of bioconversion Cassava – Δ in bioconversion from 12:1 – 6:1 in the optimistic scenario Colombia Δ from 12:1 – 6:1 Absolute ΔDALYs gained 132 % increase in DALYs gained 90 Absolute Δmonetary gain 0 % increase in monetary gain 175 Absolute ΔIRR 13 % increase in IRR 30 Honduras Δ from 12:1 – 6:1 Absolute ΔDALYs gained 328 % increase in DALYs gained 82 Absolute Δmonetary gain 0 % increase in monetary gain 139 Absolute ΔIRR 12 % increase in IRR 26 Mexico Δ from 12:1 – 6:1 Absolute ΔDALYs gained 140 % increase in DALYs gained 100 Absolute Δmonetary gain 1 % increase in monetary gain 122 Absolute ΔIRR 17 % increase in IRR 24 Nicaragua Δ from 12:1 – 6:1 Absolute ΔDALYs gained 21 % increase in DALYs gained 1 Absolute Δmonetary gain 0 % increase in monetary gain 1 Absolute ΔIRR 0 % increase in IRR 0 85 Maize – Δ in bioconversion from 12:1 – 6:1, in the optimistic scenario Colombia Δ from 12:1 - 6:1 Absolute ΔDALYs gained 221 % increase in DALYs gained 85 Absolute Δmonetary gain 1 % increase in monetary gain 96 Absolute ΔIRR 14 % increase in IRR 25 Honduras Δ from 12:1 - 6:1 Absolute ΔDALYs gained 345 % increase in DALYs gained 14 Absolute Δmonetary gain 0 % increase in monetary gain 15 Absolute ΔIRR 4 % increase in IRR 4 Mexico Δ from 12:1 - 6:1 Absolute ΔDALYs gained 7767 % increase in DALYs gained 75 Absolute Δmonetary gain 61 % increase in monetary gain 76 Absolute ΔIRR 27 % increase in IRR 13 Nicaragua Δ from 12:1 - 6:1 Absolute ΔDALYs gained 24 % increase in DALYs gained 1 Absolute Δmonetary gain 0 % increase in monetary gain 1 Absolute ΔIRR 0 % increase in IRR 0 86 Maize – Δ in bioconversion from 12:1 – 24:1, in both scenarios Colombia Δ from 12:1 - 24:1 Optimistic Pessimistic Absolute ΔDALYs gained 124 16 % decrease in DALYs gained 48 52 Absolute Δmonetary gain 0 0 % decrease in monetary gain 67 0 Absolute ΔIRR 13 9 % decrease in IRR 24 54 Δ from 12:1 - 24:1 Honduras Optimistic Pessimistic Absolute ΔDALYs gained 735 205 % decrease in DALYs gained 31 45 Absolute Δmonetary gain 1 0 % decrease in monetary gain 32 61 Absolute ΔIRR 9 11 % decrease in IRR 11 23 Δ from 12:1 - 24:1 Mexico Optimistic Pessimistic Absolute ΔDALYs gained 4793 624 % decrease in DALYs gained 47 49 Absolute Δmonetary gain 38 5 % decrease in monetary gain 47 50 Absolute ΔIRR 28 22 % decrease in IRR 13 16 Nicaragua Δ from 12:1 - 24:1 Optimistic Pessimistic Absolute ΔDALYs gained 92 326 % decrease in DALYs gained 4 43 Absolute Δmonetary gain 0 0 % decrease in monetary gain 4 53 Absolute ΔIRR 0 11 % decrease in IRR 0 20 87 Appendix I: Sensitivity to income levels Case with cassava – Δ in income levels to USD 1,000, in both scenarios Variation in IRR per year, and B/C when GNP is changed from USD 2,470 to 1,000 in the case of introducing biofortified cassava Colombia Optimistic Pessimistic Absolute ΔIRR 17 - % decrease in IRR 41 - Absolute ΔB/C 9 0 % decrease in B/C 63 57 Variation in IRR per year, and B/C when GNP is changed from USD 1,200 to 1,000 in the case of introducing biofortified cassava Honduras Optimistic Pessimistic Absolute ΔIRR 4 2 % decrease in IRR 8 154 Absolute ΔB/C 3 0 % decrease in B/C 17 22 Variation in IRR per year, and B/C when GNP is changed from USD 7,870 to 1,000 in the case of introducing biofortified cassava Mexico Optimistic Pessimistic Absolute ΔIRR 41 - % decrease in IRR 58 - Absolute ΔB/C 48 2 % decrease in B/C 87 88 Case with cassava in Nicaragua – Δ in income levels to USD 500, in both scenarios Variation in IRR per year, and B/C when GNP is changed from USD 1,000 to 500 in the case of introducing biofortified cassava Nicaragua Optimistic Pessimistic Absolute ΔIRR 16 11 % decrease in IRR 21 33 Absolute ΔB/C 32 4 % decrease in B/C 50 50 88 Case with maize – Δ in income levels to USD 1,000, in both scenarios Variation in IRR per year, and B/C when GNP is changed from USD 2,470 to 1,000 in the case of introducing biofortified maize Colombia Optimistic Pessimistic Absolute ΔIRR 19 13 % decrease in IRR 35 1 Absolute ΔB/C 16 2 % decrease in B/C 63 63 Variation in IRR per year, and B/C when GNP is changed from USD 1,200 to 1,000 in the case of introducing biofortified maize Honduras Optimistic Pessimistic Absolute ΔIRR 5 4 % decrease in IRR 5 7 Absolute ΔB/C 17 3 % decrease in B/C 17 17 Variation in IRR per year, and B/C when GNP is changed from USD 7,870 to 1,000 in the case of introducing biofortified maize Mexico Optimistic Pessimistic Absolute ΔIRR 82 59 % decrease in IRR 38 44 Absolute ΔB/C 3496 431 % decrease in B/C 87 87 Case with cassava in Nicaragua – Δ in income levels to USD 500, in both scenarios Variation in IRR per year, and B/C when GNP is changed from USD 1,000 to 500 in the case of introducing biofortified maize Nicaragua Optimistic Pessimistic Absolute ΔIRR 17 14 % decrease in IRR 20 25 Absolute ΔB/C 41 14 % decrease in B/C 50 50 89 Appendix J: Sensitivity of dietary VA intake of women 19-49 years in Colombia Cassava – Δ in VA-intake in the Colombian women of reproductive age: 636, 646, 656 µg/day Calculated gains of DALYs, USD and reduction of the health burden in optimistic and pessimistic scenarios if pro-V A- biofortified cassava would be introduced. Scenario Health burden DALYs gained Monetary gain reduction (%) (years) (million US$) IRR B/C Optimistic 0.80% 147 0.2 42.0 13.7 Pessimistic 0.00% 7 -0.2 -1.6 0.7 Calculated gains of DALYs, USD and reduction of the health burden in optimistic and pessimistic scenarios if pro-V A- biofortified maize would be introduced. Scenario Health burden DALYs gained Monetary gain reduction (%) (years) (million US$) IRR B/C Optimistic 0.70% 141 0.2 42.7 14.2 Pessimistic 0.00% 7 -0.2 -1.1 0.7 Calculated gains of DALYs, USD and reduction of the health burden in optimistic and pessimistic scenarios if pro-V A- biofortified maize would be introduced. Scenario Health burden DALYs gained Monetary gain reduction (%) (years) (million US$) IRR B/C Optimistic 0.80% 154 0.2 43.6 14.9 Pessimistic 0.00% 8 -0.2 -0.6 0.7 90