Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania ILRI PROJECT REPORT ISBN: 92–9146–695–6 The International Livestock Research Institute (ILRI) works to improve food and nutritional security and reduce poverty in developing countries through research for efficient, safe and sustainable use of livestock. Co-hosted by Kenya and Ethiopia, it has regional or country offices and projects in East, South and Southeast Asia as well as Central, East, Southern and West Africa. ilri.org CGIAR is a global agricultural research partnership for a food-secure future. Its research is carried out by 15 research centres in collaboration with hundreds of partner organizations. cgiar.org Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Kumlachew Geremew and Tadelle Dessie International Livestock Research Institute (ILRI) December 2021 ©2021 International Livestock Research Institute (ILRI) ILRI thanks all donors and organizations which globally support its work through their contributions to the CGIAR Trust Fund This publication is copyrighted by the International Livestock Research Institute (ILRI). It is licensed for use under the Creative Commons Attribution 4.0 International Licence. To view this licence, visit https://creativecommons. org/licenses/by/4.0. Unless otherwise noted, you are free to share (copy and redistribute the material in any medium or format), adapt (remix, transform, and build upon the material) for any purpose, even commercially, under the following conditions: ATTRIBUTION. The work must be attributed, but not in any way that suggests endorsement by ILRI or the author(s). NOTICE: For any reuse or distribution, the licence terms of this work must be made clear to others. Any of the above conditions can be waived if permission is obtained from the copyright holder. Nothing in this licence impairs or restricts the author’s moral rights. Fair dealing and other rights are in no way affected by the above. The parts used must not misrepresent the meaning of the publication. ILRI would appreciate being sent a copy of any materials in which text, photos etc. have been used. Editing, design and layout—ILRI Editorial and Publishing Services, Addis Ababa, Ethiopia. Cover photo—ILRI/Apollo Habtamu and ILRI/J.Bruno ISBN: 92–9146–695–6 Citation: Geremew, K. and Dessie, T. 2021. Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania. Nairobi, Kenya: ILRI. Patron: Professor Peter C Doherty AC, FAA, FRS Animal scientist, Nobel Prize Laureate for Physiology or Medicine–1996 Box 30709, Nairobi 00100 Kenya ilri.org Box 5689, Addis Ababa, Ethiopia Phone +254 20 422 3000 better lives through livestock Phone+251 11 617 2000 Fax +254 20 422 3001 Fax+251 11 667 6923 Email ilri-kenya@cgiar.org ILRI is a CGIAR research centre Email ilri-ethiopia@cgiar.org ILRI has offices in East Africa • South Asia • Southeast and East Asia • Southern Africa • West Africa iv Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Contents Tables v Figures vi Executive summary vii Introduction 1 Chicken products consumed at baseline and endline, Ethiopia 3 Descriptive summary of household dietary diversity, Ethiopia 4 Impact of intervention on household dietary diversity (HDD), Ethiopia 6 Impact of intervention on household members dietary diversity, Ethiopia 7 Chicken meat and egg consumed and used in Ethiopia 8 Chicken meat and egg consumed in Tanzania 12 Chicken meat consumed in Tanzanian ACGG+ATONU project 14 D–I–D analysis of consumed chicken, Tanzania 15 References 17 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania v Tables Table 1. Summary of household dietary diversity, Ethiopia 4 Table 2. Summary of minimum dietary diversity for women and children, Ethiopia 5 Table 3. D–I–D result for HDD, Ethiopia 6 Table 4. D–I–D results for woman and child dietary diversity, Ethiopia 7 Table 5. Eggs prepared during past week, Ethiopia 9 Table 6. Who consumed eggs prepared during last week, Ethiopia? 9 Table 7. Children’s consumption of chicken and eggs in the previous 7 days and 24 hours, Ethiopia 10 Table 8. Children’s consumption of chicken in the previous 7 days and 24 hours by treatment arm, Ethiopia 10 Table 9. Children’s consumption of eggs in the previous 7 days and 24 hours by treatment arm, Ethiopia 10 Table 10. Women’s consumption of chicken and eggs in the previous 7 days and 24-hour recall, Ethiopia 11 Table 11. Women’s consumption of chicken in the previous 7 days and 24 hours by treatment arm, Ethiopia 11 Table 12. Women’s consumption of eggs in the previous 7 days and 24 hours by treatment arm, Ethiopia 11 Table 13. Household consumed chicken per annum at baseline and endline, Tanzania 14 Table 14. Parameter estimates of the adjusted linear mixed model for D–I–D analysis of local chicken consumption, Tanzania 15 Table 15. Parameter estimates of the adjusted linear mixed model for D–I–D analysis of improved chickens’ consumption, Tanzania 16 vi Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Figures Figure 1. Chicken, eggs, vitamin A rich vegetables, and other vegetables and fruit consumed in Ethiopia 3 Figure 2. Egg production and uses, last week, Ethiopia 8 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania vii Executive summary The African Chicken Genetics Gain (ACGG) Agriculture to Nutrition (ATONU) is striving to improve nutrition outcomes through optimized agricultural investments project, aimed to develop, implement, and evaluate nutrition sensitive interventions (NSIs) within the context of existing agricultural programs with the goal of improving the nutritional status of women of reproductive age and young children, particularly in the first 1,000 days of life. ACGG ATONU implemented a NSI in collaboration with the ACGG program, led by the International Livestock Research Institute (ILRI). ACGG is evaluating the productivity of high producing, tropically adapted and farmer preferred chicken genotypes in Ethiopia, Tanzania and Nigeria. ACGG has provided approximately 25 chickens to each participating chicken producing smallholder households. ACGG ATONU had implemented an additional packages of NSIs among ACGG households using social and behavioural change communication (SBCC) to encourage consumption of chicken products (meat and eggs); recommend water, sanitation and hygiene (WaSH) practices to produce poultry; use income from sale of chicken products to improve nutrition and health; empower women in decision-making around chicken production and sale; and home gardening to produce nutrient dense vegetables to improve dietary diversity within the household. It is worth mentioning that ACGG’s intervention to increase chicken production helped to improve the nutritional status of women and children through increasing access to chicken meat and eggs for household consumption and empowering women by giving them access to income, which could be used to purchase other nutrient dense foods. However, increasing production and income alone may not necessarily translate into improved diets or nutritional outcomes. ACGG ATONU’s intervention specifically encouraged the use of chicken products and income to provide nutritious diets for women of reproductive age and young children through extensive nutrition behaviour change communication. Recognizing the unavailability of other nutrient dense foods in local markets helped the project implementers to introduce the home gardening component of ACGG ATONU’s intervention that increased the availability of nutrient dense vegetables at the household level. Following the implementation of the NSI’s smallholder farmers’ consumption pattern, women and children dietary diversity, chicken meat and egg consumption and use, hygiene and sanitation conditions and women empowerment are improved. Field officers have been visiting the homes of smallholder farmers and they witnessed that these farmers are bringing positive changes towards balanced diet consumption, good hygiene and sanitation and lower workload on women. For these results to be recorded, weekly nutrition education sessions, cookery sessions, home to home visits, theatre for nutrition and behavioural change communication, joint financial planning and budgeting and male sensitization were used as approaches. Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania 1 Introduction Undernutrition remains a key challenge to global health and development. Data indicate that a quarter (25%) of the world’s children under five years are stunted, and this prevalence increases to 38% in sub-Saharan Africa (SSA) (UNICEF 2014). Stunting is a consequence of chronic nutritional deprivation that can begin during and even prior to pregnancy due to maternal malnutrition and other adversities. Child undernutrition is associated with high child mortality and morbidity (Black et al. 2013, (Olofin & McDonald, 2013), poor motor and cognitive development (Sudfeld et al. 2015), and lower educational attainment and economic productivity in adulthood (Victora, Adair, Fall, & Hallal, 2008). Maternal short stature, a long-term consequence of stunting in girls, is further associated with foetal growth restriction, which can lead to neonatal death and stunting in the next generation (Katz et al. 2013). Similarly, maternal and child anaemia has multifactorial causes involving complex interactions with nutrition, with serious consequences for maternal and child survival and health, healthy pregnancies, cognitive development, and work productivity (Balarajan et al. 2011). Agriculture investments and productivity are increasing in Africa. Despite investments in agriculture and gains in productivity of food staples, the number of children in Africa who are stunted due to malnutrition is climbing (Stevens et al. 2012). To a great extent, achievements in food security have not translated to improved nutrition and health at the household level (IFPRI 2015). The scale of malnutrition across the continent continues to be high, with 58 million children under age of five are too short for their age (stunted), 13.9 million weighs too little for their height (wasted), and 10.3 million are overweight (UNICEF, 2015). An estimated 163.6 million children and women of reproductive age are anaemic (Kumlachew, Tesfahun, Mulugeta, & Tadelle, 2020) while more than 220 million people are estimated to be calorie deficient (FAO, IFAD, WFP 2015). Adult obesity is on the rise in all 54 African countries, with about 8% of adults over 20 are obese (Kumlachew et al., 2020). Regions leading in agricultural production tend to be the most affected by high rates of child stunting (Tanzania Food and Nutrition Centre 2014). The ripple effects of malnutrition extend beyond the affected individuals to the society and the economy. Agricultural development initiatives have the potential to improve the nutrition of those most vulnerable to malnutrition, such as women of childbearing age and children as well as the general population; however, they must be specifically designed to address nutritional goals (Ruel and Alderman 2013), with rigorous evaluation of impact on nutritional outcomes. In SSA, smallholder chicken production is an integral component of the livelihoods of poor rural households, and this is likely to continue for the foreseeable future (Kryger et al. 2010). Chicken production is largely a woman’s enterprise. In addition to various sociocultural uses, the main outputs from family chicken production are meat or eggs as a source of high value nutrition for home consumption (ideal for the most vulnerable members of the household, such as pregnant women, children and the sick) and income from the sale of birds and eggs. Eggs are an important source of energy and provide 27% of an adult’s daily requirements of selenium, 25% of vitamin B12, 23% of chlorine, 15% of riboflavin, 13% of protein, 11% of phosphorus, 9% of vitamin D, 9% of foliate, 8% of vitamin A, 6% of iron, and a small amount of zinc (FAO 2015). With nearly 40% of global chicken exports going to Africa and the Middle East (Mulder and Nelson 2013), there is tremendous opportunity for income growth and increased food security in SSA through improved chicken production (Iannotti et al. 2014). Evidence suggests that increased consumption of animal 2 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania foods can improve child growth and development (Marquis et al. 1997). However, limited evidence exists linking increased access to chicken meat and eggs with increased household animal source food consumption and improved nutrition status among women of reproductive age and children (Nielsen et al. 2003; Leroy and Frongillo 2007). Likewise, evidence for integrating NSIs in agriculture projects for impact on nutrition and health is lacking. This fundamental gap in agricultural development, nutrition and health necessitates an integrated approach towards nutrition sensitive agriculture programs. As this approach evolves in developing countries, ATONU programs become more complex and thus increases the need for novel methods of implementing and measuring the impact. Integrated ATONU interventions cover a broad range of interventions from soil conservation and management to improved germplasm and other inputs, post-harvest processing and value addition, and storage, as well as marketing and food consumption. These processes collectively affect productivity and diversity of the harvest, food safety, and the energy as well as nutrient density of the diet. Specific interventions to integrate along the ATONU continuum may differ from one country to the other depending on national food and nutrition security, available resources, gender and cultural considerations, and hence are also likely to change over time. To successfully implement integrated approaches and ensure appropriate scale up for needed interventions, countries need coordinated efforts and robust evidence to inform decision-making frameworks. This project was designed to assess the impact of integrating NSIs in a targeted chicken production project (ACGG) on maternal and child nutrition and health outcomes. ACGG was identified as a project that integrates and assesses the impact of selected NSIs to provide evidence for agricultural potential to deliver positive nutritional outcomes. In designing and implementing the NSIs, ILRI partnered with two country implementing partners namely: the Ethiopian Institute of Agricultural Research and the Tanzania Livestock Research Institute. The Harvard School of Public Health and its local Ethiopian partner, Addis Continental University, were responsible for the designing and conducting of the impact evaluation for the NSIs in Ethiopia, while Sokoine University of Agriculture was responsible for the impact evaluation in Tanzania. ACGG ATONU project was built on ACGG project that works with partners and beneficiaries to design and evaluate effective agriculture tailored nutrition interventions, and advocate for them. ACGG ATONU’s primary beneficiaries were smallholder farm families in four regional states; Amhara, Tigray, Oromia and Southern Nations Nationalities and Peoples’ Region in Ethiopia and in three agricultural zones; central, southern highlands and eastern zones in Tanzania. The project targets were women of childbearing age and young children in the first 1,000 days of life in rural households, where high nutritional demands of pregnancy, development and early childhood must largely be met through food grown, or income earned, on family farms in both Ethiopia and Tanzania. The project benefited 1,600 households in both countries. Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania 3 Chicken products consumed at baseline and endline, Ethiopia One of the major focuses of ATONU intervention is enhancing poultry and vegetable product consumption through improved production and management practices. Households in the intervention areas were provided chicken to improve own consumption and sale of eggs and live birds and encouraged to produce vegetables for the same purpose. During the baseline and endline assessment data were generated on the production and consumption of these commodities. Figure 1 presents a summary of index of child eggs and vegetable consumption in the previous seven days. In general, compared to the baseline, consumption of both food items has increased at the endline in all treatment arms. However, the increase in consumption of eggs looks significantly higher in the ACGG+ATONU treatment arm than both control and ACGG treatment arms. Compared to the control and ACGG treatment arms, a higher proportion of children in the ACGG+ATONU arm households consumed eggs that could be attributed to the NSI. Figure 1. Chicken, eggs, vitamin A rich vegetables, and other vegetables and fruit consumed in Ethiopia. 4 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Descriptive summary of household dietary diversity, Ethiopia Household dietary diversity index (HDD) is one of the most important indicators used to assess the food security status of households. It helps to measure the number of different types of food and food groups household consume and shows the diversity of food groups the household consume. It can be measured either at households or individual level such as women or children. Even though there are no standard cut off points that would indicate the adequacy of measured values, some institutions like IFPRI have their cut off points to assess the adequacy level. Table 1 presents a summary of household dietary diversity indicators by intervention period using 12 food groups approach at the household level and 9 food groups approach for women level and 7 food groups approach for child level. The household level dietary diversity is measured using 30 days food consumption recall information and usually, it does not only indicate the food security status of households, but also the wealth of households. As a result, it includes more food groups that are not included at individual level measurements. The overall average household dietary diversity score was about 4.13 during baseline and increased to 4.64 at the endline. The test for the presence of a difference between the mean values at different periods shows the presence of a statistically significant difference during the three periods. Table 1. Summary of household dietary diversity, Ethiopia Baseline Midline Endline Total Food security indicator Chi-squared Mean SD Mean SD Mean SD Mean SD HHDD_30DY 4.13 1.76 4.41 1.76 4.64 1.85 4.39 1.80 76.7*** WDD_7NEW 3.70 1.56 4.14 1.61 4.27 1.67 4.03 1.63 152.4*** WDD_YNEW 2.72 1.06 2.93 1.16 3.14 1.20 2.93 1.15 132.2*** CHDD_7NEW 2.14 1.71 2.39 1.69 2.68 1.85 2.42 1.77 22.05*** CHDD_YNEW 1.79 1.45 2.05 1.44 2.22 1.54 2.03 1.49 22.5*** **Inference: *** p< 0.01; ** p< 0.05;*p< 0.1. Women’s dietary diversity was measured using 7 days and 24 hours of food consumption information. The 7 days consumption dietary diversity shows women’s dietary diversity grew from 3.7 food groups at baseline to 4.27 food groups at the endline. The average difference at baseline and endline seems statistically significant. Likewise, the 24 hours food consumption recall also indicates women’s food consumption diversity has increased from 2.72 to 3.14 food groups, with a statistically significant difference between the baseline and endline. Children’s dietary diversity was also measured using the 7 days and 24 hours of food consumption recall. On average the seven days consumption of child dietary diversity grew from 2.14 during baseline to 2.68 during the endline. Similarly, the 24 hours recall consumption also grew from 1.79 food groups to 2.22 food groups. The test for difference in the average values suggests the presence of a statistically significant difference between baseline and endline. Minimum dietary diversity is also the other binary food security indicator that indicates the food security status of households, women, and children. This indicator is constructed using certain cut off points, where individuals above the cut-off point are considered as a food secure and below the cut-off point as otherwise. Table 2 presents Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania 5 a summary of women and child level indicators measured in 7 days and 24 hours recall. Based on the 7 days recall measure, the proportion of women that fulfil the minimum dietary diversity increased from 27.9 to 43.0%, which seems very significant. Similarly, based on the 24 hours recall measure, the proportion of women considered as an inadequate dietary diversity increased from 4.9 to 11.8%. The test for the presence of an association between minimum dietary diversity and period of observation is statistically significant at 1%, suggesting the presence of a strong association between periods of observation and level of minimum dietary diversity. Like the women minimum dietary diversity indicator, the proportion of children who were considered inadequate dietary diversity increased from 22.2 to 35.5%. There is also a statistically significant association between children’s minimum dietary diversity and period of observation. In general, the minimum dietary diversity indicator of women and children suggests the presence of significant improvement during the endline than the baseline, which could be associated with the project interventions or other changes outside the project intervention. Table 2. Summary of minimum dietary diversity for women and children, Ethiopia Minimum dietary diversity Baseline Midline Endline Total Pearson chi2 HMDD Yes 21.6 26.4 32.8 26.9 No 78.4 73.6 67.2 73.1 66.6*** WMDD7 Yes 27.9 37.4 43.0 36.0 No 72.1 62.6 57.0 64.0 105.5*** WMDDY Yes 4.9 9.1 11.8 8.6 No 95.1 90.9 88.2 91.4 64.2*** CMDD7 Yes 22.2 26.2 35.5 28.2 No 77.8 73.8 64.5 71.8 45.9*** CMDDY Yes 12.3 15.4 20.7 16.3 No 87.7 84.6 79.3 83.7 26.5*** **Inference: *** p< 0.01; ** p<0.05;*p<0.1. 6 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Impact of intervention on household dietary diversity (HDD), Ethiopia The impact of the project intervention on household food security indicators was measured using the most commonly used difference-in-difference (D–I–D) approaches. Table 3 presents estimated results on the HDD indicators for the three treatment arms (Control vs ACGG+ATONU; Control vs ACGG; ACGG vs ACGG+ATONU). The impact of ACGG+ATONU interventions on HDD seems statistically significant at 1%. Compared to households in the control arms, households in the ACGG+ATONU had higher HDD, which could be attributed to the project intervention. Compared to the control arm, there was also a positive change in the ACGG treatment arm though the difference is not statistically significant. Furthermore, considering ACGG as the control group, the impact of ACGG+ATONU intervention on HDD seems positive despite the absence of a statistically significant difference between the two groups. This shows that in addition to the chicken-based intervention the NSI in the ATONU villages brought some change in HDD. Table 3. D–I–D result for HDD, Ethiopia Before After D–I–D (T–C) Treatment arms CT TR Diff (T–C) T CT TR Diff (T–C) T Diff (T–C) T Control vs ACGG+ATONU 0.253 0.232 –0.021 –0.88 0.295 0.375 0.08 3.27*** 0.101 2.96*** Control vs ACGG 0.238 0.218 –0.019 –0.82 0.3 0.325 0.025 1.05 0.045 1.33 ACGG vs ACGG+ATONU 0.224 0.232 0.007 0.3 0.332 0.375 0.043 1.76* 0.036 1.05 **Inference: *** p< 0.01; ** p<0.05;*p<0.1. Where CT=Control, TR=Treatment, T–C=Treatment minus Control, T=Treatment Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania 7 Impact of intervention on household members dietary diversity, Ethiopia In addition to household level aggregated data, intra household level indicators can also give us a good picture of the nutritional status of the selected household members who are vulnerable to food insecurity. Using the D–I–D approaches, the impact of the two interventions on women and children’s dietary diversity is presented in Table 4. The estimated result suggests the impact of ACGG+ATONU intervention on women’s dietary diversity measured at seven days recall seem statistically significant. Compared to women in the control arm, women in the ACGG+ATONU arms had higher dietary diversity at the endline than at the baseline. This could be attributed to the NSIs in the ATONU treatment arm. The D–I–D test between the ACGG and ACGG+ATONU intervention also shows the presence of a positive change in the ACGG+ATONU arms than the ACGG arms. However, this difference is not statistically significant. The positive difference in the values of these two intervention arms is expected due to the impact of other nutrition-based interventions in addition to the chicken related intervention. Table 4. D–I–D results for woman and child dietary diversity, Ethiopia Before After D–I–D (T–C) Indicator Treatment arms Diff Diff Diff CT TR T CT TR T T (T–C) (T–C) (T–C) Woman Control vs 1.56 2.08*** dietary ACGG+ATONU 3.694 3.825 0.131 4.134 4.513 0.38 4.43*** 0.249 diversity –7D Control vs ACGG 3.665 3.758 0.093 1.12 4.116 4.299 0.183 2. 14*** 0.09 0.75 ACGG vs 0.44 2.03*** 1.14 ACGG+ATONU 3.786 3.825 0.039 4.332 4.513 0.181 0.143 Child dietary Control vs –2.02*** 1.92** 0.146 2.97*** diversity–7D ACGG+ATONU 0.305 0.229 –0.076 0.362 0.432 0.07 Control vs ACGG 0.246 0.229 –0.016 –0.43 0.322 0.432 0.11 3.0*** 0.126 2.4*** ACGG vs –2.54*** 0.05 0.095 1.86* ACGG+ATONU 0.338 0.245 –0.094 0.311 0.313 0.002 **Inference: *** p<0.01; ** p<0.05;*p<0.1. Where CT=Control, TR=Treatment, T–C=Treatment minus Control, T=Treatment Like women’s dietary diversity, the NSIs have brought significant change in the diversity of children’s diets. Compared to the control arms, both ACGG and ACGG+ATONU interventions have brought statistically significant changes in child dietary diversity. Furthermore, the test for the relative difference between ACGG and ACGG+ATONU suggests the presence of statistical difference (10%) in dietary diversity between the two intervention areas. The above findings show if properly implemented NSIs could potentially bring positive change in food and nutrition security of rural households. 8 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Chicken meat and egg consumed and used in Ethiopia To assess the pathway between chicken production and consumption, women were specifically asked about egg consumption and cooking methods. Figure 2 describes eggs produced, consumed, sold, given away and retained for chick production. The survey questions on which this section is based ask specifically about the previous seven days. Looking at endline data, we see that the ACGG and ACGG+ATONU arms are producing approximately twice as many eggs at endline as the control arm, and more eggs than at baseline. At endline, the average number of eggs produced in the last week was 9.9, 8.5 and 4.5 for ACGG, ACGG+ATONU and control arms, respectively. We also see that in each arm, similar numbers of eggs are consumed at home as are sold. Figure 2. Egg production and uses, last week, Ethiopia. Egg Production and Uses, Last Week 10 9 8 7 6 5 4 3 2 1 0 ACGG ACGG + Control ACGG ACGG + Control ACGG ACGG + Control ATONU ATONU ATONU Baseline Midl ine Endline # eggs produced # of these sold # given away # retained for chick production # consumed Overall, while household consumption of eggs produced on-farm was low and not different among arms at baseline (p>0.05), the ACGG and ACGG+ATONU arms had higher household egg consumption (approximately 4 eggs/week) than the control arm (approximately 2 eggs/week) at endline (p<0.05). Table 5 also shows the most common egg preparation methods. On average, the ACGG+ATONU group prepared the greatest number of eggs at endline during the last preparation (6.9), which had increased from the baseline amount by about one egg (from 5.6); the ACGG arm consumed about 6.6 eggs during the last preparation, which also increased from 5.6 at baseline. The control arm prepared about 5.8 eggs during the last preparation, which also saw an increase from about 5.1 at baseline. The most common cooking method was frying and scrambling for all treatment arms. There seemed to be a slight decrease in boiling eggs, and a slight increase in preparation by cooking in a stew or meat dish compared to baseline for all treatment arms. Number of eggs Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania 9 Table 5. Eggs prepared during past week, Ethiopia Baseline Midline Endline Mean Mean Mean Mean Mean Mean Mean Mean Mean Prepared egg(s) in past week 27% 32% 27% 37% 41% 28% 46% 51% 28% No. of eggs prepared 5.6 5.8 5.1 6.5 6.7 5.6 6.6 6.9 5.8 How were they prepared? % Boiled and combined in a meat dish 2% 4% 2% 6% 9% 5% 12% 7% 7% Boiled (not combined in a meat dish) 20% 18% 13% 14% 12% 16% 11% 12% 8% Fried, scrambled 56% 50% 43% 44% 42% 33% 53% 51% 46% Fried, whole 11% 11% 15% 14% 12% 16% 7% 11% 13% Cooked in stew 12% 18% 25% 21% 23% 29% 16% 20% 27% Table 6 is based on the section of the survey that first asks about whether eggs were consumed in the past week, and if they were, how many were consumed by the following groups: the woman of reproductive age, the index child, the oldest child, other household members, the household head and non-household members. The average number of eggs consumed at each meal for the woman of reproductive age and the index child is relatively consistent from baseline to endline and across groups. However, as shown in Table 5 above, there was a much higher percentage of households that prepared eggs in the past week in the ACGG+ATONU (51%) and ACGG (46%) arms compared to the control arm (28%), meaning a greater number of children and women of reproductive age actually ate eggs in these groups, and that a greater number of eggs was consumed by these intervention groups relative to the control. Table 6. Who consumed eggs prepared during last week, Ethiopia? No. of No. of eggs No. of eggs No. of eggs other No. of eggs non No. of eggs hh eggs WRA index child oldest child household (hh) hh members head consumed consumed consumed consumed members consumed consumed Mean 1.8 2.0 1.6 3.2 2.2 0.6 SD 1.1 1.4 0.9 2.5 1.2 1.0 Mean 2.1 2.4 2.1 3.0 2.1 0.8 SD 2.1 2.2 2.4 2.1 1.5 1.4 Mean 2.0 2.3 2.1 3.1 2.3 0.7 SD 1.9 2.6 2.3 2.8 1.9 1.3 Mean 1.8 1.9 2.1 3.4 2.0 0.4 SD 1.1 1.5 2.1 2.3 1.2 1.0 Mean 1.8 2.0 2.0 3.5 2.1 0.7 SD 1.5 2.0 2.2 2.9 2.1 1.3 Mean 1.7 2.1 1.6 3.1 2.0 0.4 SD 1.2 1.6 1.0 2.4 1.2 0.6 Mean 1.8 1.8 1.8 3.3 2.0 2.8 SD 1.3 1.2 1.6 2.9 1.6 1.9 Mean 1.8 1.7 1.6 3.9 2.0 3.4 SD 1.6 1.5 1.6 2.9 1.3 3.0 Mean 1.8 1.7 1.6 2.8 1.8 2.3 SD 1.4 1.6 2.0 2.2 1.2 1.6 WRA=Women of Reproductive Age Endline Midline Baseline ACGG+ ACGG+ ACGG+ Control ACGG Control ACGG Control ACGG ATONU ATONU ATONU ACGG ACGG+ ATONU Control ACGG ACGG+ ATONU Control ACGG ACGG+ ATONU Control 10 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Children’s consumption of chicken remained low across all arms, while egg consumption by children increased across time for all arms. Results from the 24-hour recall indicate that chicken consumed by children was low throughout the study with <1% of children consuming chicken in the previous 24 hours (Table 7). In the 7-day food frequency questionnaire, chicken consumption increased marginally to 2.7% at endline from 1% at baseline. Egg consumption by children was more frequent, with 15.4% of the children consuming eggs in the previous 24 hours at endline compared to 12% at baseline. In the previous 7 days, 27.5% of children had consumed eggs at endline compared to 24.2% at baseline. Eggs consumed appeared to be greater for the ACGG+ATONU arm at endline compared to other arms (Table 8). Table 7. Children’s consumption of chicken and eggs in the previous 7 days and 24 hours, Ethiopia Chicken consumption Egg consumption 7-day FFQ• 24 hr recall 7-day FFQ 24 hr recall No. 7 11 27 4 1 4 177 207 272 88 110 152 % 1.0 1.2 2.7 0.6 0.1 0.4 24.21 23.05 27.5 12.04 12.25 15.37 No. 731 898 989 731 898 989 731 898 989 731 898 989 •FFQ=Food Frequency Questionnaire. Table 8. Children’s consumption of chicken in the previous 7 days and 24 hours by treatment arm, Ethiopia Chicken consumption 7-day FFQ• Chicken consumption 24-hour recall Baseline Midline Endline Baseline Midline Endline No. 4 1 2 4 5 2 7 8 12 0 0 4 0 0 1 0 3 1 % 1.7 0.5 0.8 1.4 1.8 0.6 2.3 2.6 3.2 0.0 0.0 1.5 0.0 0.0 0.3 0.0 1.0 0.3 No. 243 222 266 286 278 334 305 307 377 243 222 266 286 278 334 305 307 377 •FFQ=Food Frequency Questionnaire. Based on the FFQ table, eggs consumed in the ACGG+ATONU arm for the last 7 days was improved from 27% at baseline to 35.8% at endline among children. This improvement is registered as a significant change resulted from the ATONU packages (Table 9). Taking the 24-hour recall method into account, egg consumption among indexed children of the ACGG+ATONU treatment arm showed a significant increase from 15.4% at baseline to 21.8% at endline (Table 9). Table 9. Children’s consumption of eggs in the previous 7 days and 24 hours by treatment arm, Ethiopia Egg consumption 7-day FFQ• Egg consumption 24-hr recall Baseline Midline Endline Baseline Midline Endline No. 58 60 59 76 73 58 82 110 80 26 31 31 44 39 27 47 67 38 % 23.9 27.0 22.2 26.6 26.3 17.4 26.9 35.8 21.2 10.7 14.0 11.7 15.4 14.0 8.1 15.4 21.8 10.1 No. 243 222 266 286 278 334 305 307 377 243 222 266 286 278 334 305 307 377 •FFQ=Food Frequency Questionnaire. ACGG ACGG+ ACGG Baseline ATONU ACGG+ Control ATONU Midline ACGG Control ACGG+ Endline ATONU ACGG Control ACGG+ Baseline ATONU ACGG Control Midline ACGG+ ATONU ACGG Endline Control ACGG+ ATONU ACGG Control Baseline ACGG+ ATONU ACGG Midline Control ACGG+ ATONU ACGG Control Endline ACGG+ ATONU ACGG ACGG+ Baseline Control ATONU ACGG Control Midline ACGG ACGG+ ATONU ACGG+ ATONU Endline Control Control Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania 11 On average, 11.5% of women reported consuming eggs in the 24-hour dietary recall at endline, compared to 6.9% at midline and 4.6% at baseline. Comparatively, 32.8% of the women reported consuming eggs on at least one occasion in the past 7 days, compared to 25.4% at midline and 19.4% at baseline. At endline, 1% of women reported consuming chicken in the previous 24 hours compared to 0.5% at baseline; 5.4% ate chicken on at least one occasion in the past week compared to 2.2% at baseline (Table 10). Chicken consumption remained very low throughout the study, while egg consumption showed marginal increases in all arms. Eggs consumed appeared to be greater for the ACGG and ACGG+ATONU arms at endline compared to the control arm (Table12). Table 10. Women’s consumption of chicken and eggs in the previous 7 days and 24-hour recall, Ethiopia Consumed chicken Consumed eggs 7-day FFQ• 24-hr recall 7-day FFQ 24-hr recall No. 46 87 110 11 10 20 410 518 664 98 141 233 % 2.2 4.3 5.4 0.5 0.5 1.0 19.4 25.4 32.8 4.63 6.91 11.49 No. 2,117 2,041 2,027 2,117 2,041 2,027 2,117 2,041 2,027 2,117 2,041 2,027 •FFQ=Food Frequency Questionnaire. Based on the FFQ table chicken meat consumed in the ACGG+ATONU arm for the last 7 days was improved from 1.8% at baseline to 5.1% at endline among women. This improvement is registered as a significant change resulted from the ATONU packages (Table 11). Taking the 24-hour recall method into account, chicken meat consumed among women of reproductive age of the ACGG+ATONU treatment arm also showed a slight increase from 0.4% at baseline to 0.7% at endline (Table 11). Table 11. Women’s consumption of chicken in the previous 7 days and 24 hours by treatment arm, Ethiopia Consumed chicken in 7-day FFQ• Consumed chicken in 24-hour recall Baseline Midline Endline Baseline Midline Endline No, 20 13 13 34 35 18 41 43 26 5 3 3 2 7 1 8 5 7 % 2.8 1.8 1.9 5.0 5.1 2.7 6.1 6.3 3.9 0.7 0.4 0.4 0.3 1.0 0.2 1.2 0.7 1.0 No, 710 709 698 684 687 670 673 681 673 710 709 698 684 687 670 673 681 673 Table 12. Women’s consumption of eggs in the previous 7 days and 24 hours by treatment arm, Ethiopia Consumed egg in 7-day FFQ• Consumed egg in 24-hour recall Baseline Midline Endline Baseline Midline Endline No, 122 160 128 184 211 123 238 278 148 23 44 31 46 66 29 83 98 52 % 17.2 22.6 18.3 26.9 30.7 18.4 35.4 40.8 22.0 3.2 6.2 4.4 6.7 9.6 4.3 12.3 14.4 7.7 No, 710 709 698 684 687 670 673 681 673 710 709 698 684 687 670 673 681 673 •FFQ=Food Frequency Questionnaire. ACGG ACGG ACGG+ ATONU ACGG+ Baseline ATONU Control Control ACGG Midline ACGG ACGG+ ATONU ACGG+ Endline Control ATONU Control ACGG Baseline ACGG ACGG+ ATONU Midline ACGG+ Control ATONU ACGG Control Endline ACGG+ ACGG ATONU Baseline ACGG+ Control ATONU Control Midline ACGG ACGG ACGG+ Endline ATONU ACGG+ ATONU Control Baseline Control ACGG ACGG Midline ACGG+ ATONU ACGG+ ATONU Control Endline Control 12 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Chicken meat and egg consumed in Tanzania The Agriculture to Nutrition (ATONU) is trying to improve nutrition outcomes through optimized agricultural investments project sought to break the intergenerational cycle of undernutrition and answer the question of what agriculture can do to deliver positive nutrition outcomes through tailored NSIs targeting women of childbearing age and children in the first 1,000 days of life, that is, from conception to two years. ATONU has developed frameworks to select and provide technical support to agricultural projects for integrating NSIs, identifying, selecting, and implementing NSIs and evaluating their impact. The project’s approach was to work with existing agriculture development projects to select and implement tailor made interventions, evaluate their impact, and provide evidence of what agriculture can do to deliver positive nutrition outcomes. ATONU selected the ACGG project, which was implemented by ILRI in several African countries, including Ethiopia and Tanzania. The ACGG project’s aim was to improve the production and productivity of chickens kept by smallholder households by introducing and testing the performance of improved and tropically adapted genotypes. Although ACGG was implemented in five zones, ATONU interventions were implemented in three zones of Tanzania, namely eastern, central and southern highlands. The following interventions were selected and overlaid onto the ACGG project and delivered as a package to participating households: • Social and behavioural changes communication (SBCC) on nutrition education and hygiene to increase eggs and chicken meat consumed. • SBCC to increase expenditure on nutritious food through income generated from sale of eggs and chicken. • SBCC to empower women to influence changes in women’s time use and participation in decision-making within the household. • Promoting home gardens to increase vegetable production for improving dietary diversity. This study hypothesized that both ACGG and ATONU can improve women’s and children’s diets via the following three pathways: food production for own consumption, increased income expenditure on additional nutrient dense foods, and women’s empowerment, all working in concert. Specifically, ACGG could increase chicken meat and eggs produced, which would increase access to these products for household consumption. These products could also be marketed, providing a source of income that could be used in part for improving diets. ACGG’s efforts to target women could also lead to greater engagement and participation by women in household decision-making on chicken production and marketing, which could, in turn, empower them and improve their status within the household. The SBCC on nutrition education and hygiene could encourage household consumption of chicken products and other nutrient dense foods, especially by women and children. SBCC on household budgeting could encourage use of income from chicken production, specifically for the purchase of other foods that could not be produced by the household but provide nutrients that would be missing in the household diet. Interventions addressing gender dynamics within the household could further empower women in chicken production and other aspects of household life. Household members’ adoption of appropriate WaSH behaviours could decrease harmful exposure to poultry droppings, thereby decreasing morbidity among children and improving Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania 13 food and nutrient use. Given that chicken products provide a good source of animal protein and essential amino acids and micronutrients, it was expected that infants and young children would have better growth, women would be less likely to be underweight, and both women and children would have a reduced risk of anaemia. 14 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Chicken meat consumed in Tanzanian ACGG+ATONU project Chicken meat consumed was higher at baseline than at endline (Table 13). At baseline, the mean number of chickens consumed per year ranged from 20 to 22 chickens in all treatment arms and zones. At endline, the mean number of chickens consumed was 18 (95% CI, 15.00–20.98) for ACGG+ATONU treatment arm; 20 (95% CI, 15.58–23.96) for ACGG; and 13 (95% CI, 9.69–16.78) for the control treatment arm. The extent of reduction of the number of chickens consumed per year was 7, 3 and 4 in the control, ACGG and ACGG+ATONU treatment arms, respectively. Similarly, the reduction was 6, 4 and 2 for the central, southern highlands and eastern zones, respectively. Table 13. Household consumed chicken per annum at baseline and endline, Tanzania Baseline Endline Treatment/zone No, Mean SE 95% CI No, Mean SE 95% CI Control 458 20.04 1.31 17.43–22.65 386 13.23 1.77 9.69–16.78 ACGG 453 22.48 1.77 18.94–26.02 446 19.77 2.09 15.58–23.96 ACGG+ATONU 481 21.9 2.09 17.70–26.08 512 17.99 1.49 15.00–20.98 Zone Central 564 20.04 1.55 16.94–23.13 562 14.44 0.68 13.09–15.79 Eastern 287 20.15 1.33 17.49–22.80 240 18.45 2.42 13.61–23.30 Southern highlands 541 23.68 1.88 19.92–27.44 542 19.54 2.22 15.10–23.98 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania 15 D–I–D analysis of consumed chicken, Tanzania To assess the effect of the treatment arms on change in household chicken consumption from baseline to endline, D–I–D analysis was performed using linear mixed model. Results of the fitted model are presented in Table 14 (local chickens) and Table 15 (improved chickens). The analysis was adjusted for zone and endline household wealth index. For local chickens, the results revealed that consumption increased from baseline to endline. The magnitude of the D–I–D coefficient for ACGG+ATONU vs control treatment arms was 1.1929, with p-value of 0.0473. This means that the change in local chicken consumption from baseline to endline was significantly higher in ACGG+ATONU than in the control treatment arm. However, no significant difference was observed for ACGG only compared to control treatment arm (β=1.1489, p=0.0595) and ACGG+ATONU as compared to ACGG only (β=0.04400, p=0.9407). Table 14. Parameter estimates of the adjusted linear mixed model for D–I–D analysis of local chicken consumption, Tanzania Effect Estimate Standard error P-value Intercept 2.5215 0.3924 < 0.0001 Time Endline 0.5741 0.4379 0.19 Baseline Reference Treatment ACGG –0.4452 0.4129 0.2811 ACGG+ATONU –0.5047 0.4065 0.2145 Control Reference Time*treatment 0.0853 Time*ACGG 1.1489 0.6093 0.0595 Time*ACGG+ATONU 1.1929 0.6011 0.0473 Agro-ecological zone 0.0028 Central 0.5157 0.2741 0.06 Eastern 1.1553 0.3395 0.0007 Southern highlands Reference Endline wealth index <0.0001 Lowest Reference Second 0.5326 0.3928 0.1753 Middle 0.7423 0.3972 0.0618 Fourth 1.2249 0.3997 0.0022 Highest 2.2261 0.4001 <0.0001 D–I–D coefficients Effect Estimate Standard error P-value ACGG vs control 1.1489 0.6093 0.0595 ACGG+ATONU vs control 1.1929 0.6011 0.0473 ACGG+ATONU vs ACGG 0.04400 0.5911 0.9407 16 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania Table 15. Parameter estimates of the adjusted linear mixed model for D–I–D analysis of improved chickens’ consumption, Tanzania Effect Estimate Standard error P-value Intercept 0.02877 0.2023 0.8874 Time Endline 3.5778 0.3932 <0.0001 Baseline Reference Treatment ACGG 0.004499 0.1945 0.9815 ACGG+ATONU 0.02961 0.1915 0.8771 Control Reference Time*treatment 0.0225 Time*ACGG 0.9125 0.4702 0.0525 Time*ACGG+ATONU 1.2863 0.4663 0.0059 Agro-ecological zone 0.0718 Central –0.1402 0.1578 0.3743 Eastern 0.2854 0.1894 0.132 Southern highlands Reference Endline wealth index 0.0329 Lowest Reference Second 0.006073 0.2214 0.9781 Middle 0.2681 0.2203 0.2238 Fourth 0.2817 0.2208 0.2021 Highest 0.5942 0.2186 0.0066 D–I–D coefficients Effect Estimate Standard error P value ACGG vs control 0.9125 0.4702 0.0525 ACGG+ATONU vs control 1.2863 0.4663 0.0059 ACGG+ATONU vs ACGG 0.3738 0.3612 0.3008 For improved chicken, the results of the adjusted linear mixed model presented in Table 15 showed that the number of improved chickens consumed was also higher in endline as compared to baseline. In addition, the change for improved chicken consumption was significantly higher among subjects in ACGG+ATONU (β=1.2863, p=0.0059) and ACGG only (β=0.9125, p=0.0525) in comparison to subjects in control arms. But no significant difference in change for improved chicken consumption was observed between ACGG+ATONU and ACGG only (β=0.3738, p=0.3008). Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania 17 References Balarajan, Y., Ramakrishnan, U., Özaltin, E., Shankar, A.H. and Subramanian, S.V. 2011. Anaemia in low-income and middle-income countries. The Lancet 378: 2123–2135. Black, R., Victora, C.G., Walker, S.P., Bhutta, Z.A., Christian, P. et al. 2013. Maternal and child undernutrition and overweight in low-income and middle-income countries. The Lancet 382: 427–451. FAO (Food and Agriculture Organization of the United Nations). 2015. Egg facts. Rome, Italy: FAO. FAO (Food and Agriculture Organization of the United Nations), IFAD (International Fund for Agricultural Development) and WFP (United Nations World Food Programme). 2015. The state of food insecurity in the world 2015: Meeting the 2015 international hunger targets: Taking stock of uneven progress. Rome, Italy: FAO. Iannotti, L.L., Lutter, C.K., Bunn, D.A. and Stewart, C.P. 2014. Eggs: The uncracked potential for improving maternal and young child nutrition among the world’s poor. Nutrition Reviews 72: 355–368. doi:10.1111/nure.12107. IFPRI (International Food Policy Research Institute). 2015. Global nutrition report 2015: Africa brief: Actions and accountability to advance nutrition and sustainable development. Washington, DC, USA: IFPRI. Katz, J., Lee, A.C., Kozuki, N., Lawn, J.E. Cousens, S. et al. 2013. Mortality risk in preterm and small-for-gestational-age infants in low-income and middle-income countries: A pooled country analysis. The Lancet 382: 417–425. Kryger, K.N., Thomsen, K.A., Whyte, M.A. and Dissing, M. 2010. Smallholder poultry production—Livelihoods, food security and sociocultural significance. Rome, Italy: Food FAO. Kumlachew, G., Tesfahun, A., Mulugeta, Y., and Tadelle, D. 2020. Overview on the impacts of nutrition-sensitive intervention on dietary diversity and nutritional status of women and children in developing countries. Ethiopian Journal of Public Health and Nutrition 4. Leroy, J.L. and Frongillo, E.A. 2007. Can interventions to promote animal production ameliorate undernutrition? The Journal of Nutrition 137: 2311–2316. Marquis, G.S., Habicht, J.P., Lanata, C.F., Black, R.E. and Rasmussen, K.M. 1997. Breast milk or animal product foods improve linear growth of Peruvian toddlers consuming marginal diets. The American Journal of Clinical Nutrition 66: 1102–1109. Mulder, N.D. and Nelson, D.C. 2013. Rabobank poultry quarterly: Outlook for global and regional markets. Utrecht, the Netherlands: Rabobank. Nielsen, H., Roos, N. and Thilsted, S.H. 2003. The impact of semi scavenging poultry production on the consumption of animal source foods by women and girls in Bangladesh. The Journal of Nutrition 133: 4027S–4030S. Ruel, M.T. and Alderman, H. 2013. Maternal and child nutrition study group. Nutrition sensitive interventions and programs: How can they help to accelerate progress in improving maternal and child nutrition? The Lancet 382: 506. Stevens, G.A., Finucane, M.M., Paciorek, C.J., Flaxman, S.R., White, R.A. et al. 2012. Trends in mild, moderate, and severe stunting and underweight, and progress towards MDG 1 in 141 developing countries: A systematic analysis of population representative data. The Lancet 380: 824–834. Sudfeld, C.R., Charles, D., McCoy, G., Danaei, G., Fink, M. et al. 2015. Linear growth and child development in low- and middle-income countries: A meta-analysis. Paediatrics 135: e1266–e1275. Tanzania Food and Nutrition Centre. 2014. Tanzania national nutrition survey 2014: Final report. Dodoma, Tanzania: Ministry of Health and Social Welfare. 18 Human nutrition trials around chicken meat and egg consumed in Ethiopia and Tanzania UNICEF (United Nations Children’s Fund). 2015. Actions and accountability to advance nutrition and sustainable development. Global Nutrition Report 2015. UNICEF. 2014. The state of the world’s children 2014. New York, USA: UNICEF. Victoria, C.G., Adair. L., Fall, C. and Hallal, P.C, Martorell, R., Richter, L., et al. 2008. Maternal and child undernutrition: Consequences for adult health and human capital. The Lancet 6736(07)61692-4. https://doi. org/10.1016/S0140-6736(07)61692-4 Human nutrition trails around chicken meat and egg consumed in Ethiopia and Tanzania ILRI PROJECT REPORT ISBN: 92–9146–695–6 The International Livestock Research Institute (ILRI) works to improve food and nutritional security and reduce po verty in developing countries through research for efficient, safe and sustainable use of livestock. Co-hosted by Keny a and Ethiopia, it has regional or country offices and projects in East, South and Southeast Asia as well as Central, East, Southern and West Africa. ilri.org CGIAR is a global agricultural research partnership for a food-secure future. Its research is carried out by 15 research centres in collaboration with hundreds of partner organizations. cgiar.org