Six sub-sessions
Chair: Paul Nkegbe, University of Development Studies
Speakers and presentations
Derek Headey, International Food Policy Research Institute (IFPRI)
Re-Assessing Africa’s Health and Nutritional Outcomes through the Lens of its Diverse Farming Systems
Recording
Catherine Pfeifer, Research Center for Organic Agriculture
The Role of Livestock Ownership in Child Nutrition in East Africa
Slides/Recording
Kalle Hirvonen, IFPRI
Adoption of a Government-Led Homestead Gardening Program in Ethiopia
Recording
Oluwaseun Ariyo, University of Ibadan
Effect of Home Gardening on Food Security and Diet Quality of Households in Ijebu-Igbo, Nigeria
Recording
Marianne V. Santoso, Cornell University
A Participatory Agroecological Intervention Amongst Rural Tanzanian Farmers Increases Crop Diversity, Food Security, and Maternal Dietary Diversity: A Midline Evaluation
Recording
Farai Gwelo, FANRPAN
Agriculture to Nutrition: Promoting Vegetable Production for Improving Dietary Diversity in Ethiopia and Tanzania
Slides/Recording
Q&A
Recording
Re-Assessing Africa’s Health and Nutritional Outcomes through the Lens of its Diverse Farming Systems
Derek Headey, IFPRI
Introduction
Sub-Saharan Africa’s (SSA) reliance on rainfed agriculture and its diverse agricultural conditions merit a deeper understanding of its farming systems. Farming systems represent broadly defined groups of individual farms sharing similar resource bases, livelihood strategies, and constraints. These systems not only respond similarly to agro-climate conditions but have broadly similar development pathways and policy needs. This study analyzes trends and the current state of health and nutritional outcomes in SSA through the lens of its diverse farming systems. We aim to identify system-specific challenges and potential entry points.
Methods
Data from 24 countries in SSA are analyzed through descriptive and multivariate analysis. Geo-referenced and nationally representative health, nutrition, and socioeconomic data come from Demographic and Health Surveys (DHS). Depending on the country, one to five rounds of DHS are available covering the period since the 1990s. The source of farming systems data is Dixon (2001). Gridded landscape-level data on agro-climatic, biophysical, and economic factors (e.g., elevation and market access) is CELL5M (HarvestChoice, IFPRI, and University of Minnesota, 2017). SPAM (IFPRI and IIASA, 2016) is the source of gridded data on agricultural production. Indicators of nutritional outcomes are constructed based on anthropometric data collected from children under five years old and adult women. The incidence of various diseases (e.g., diarrhea, malaria, cough, fever) are also examined. Data from the most recently available DHS waves are used to assess the current state of health and nutrition while trends are examined for a sub-sample of countries with multiple rounds of DHS. The multivariate analysis combines ordinary least squares (for linear nutritional outcomes) and logistic regression (for dichotomous outcomes).
Findings and Interpretations
Multivariate analysis is currently underway but preliminary descriptive results suggest significant heterogeneity in the incidence malnutrition and diseases by farming systems. For example, highland perennial system is associated with the lowest incidence of moderate stunting and wasting. In addition to cereals and pulses, this system is dominated by crops that are resilient to drought and heat, such as plantain, false banana, and cassava. On the other hand, the maize-mixed system – upon which about 15% of the regional population relies for livelihood – has one of the highest incidences of under-five stunting as well as diarrhea among children aged 6-23 months. Forest-based, humid lowland tree crops, roots and tubers, and cereal-root crop mixed systems are associated with a high incidence of malaria among children 2 to 10 years old. These systems generally receive above average precipitation, a factor that is conducive for vector breeding.
Conclusions
While there is an admittedly significant degree of heterogeneity within any single farming system, identification of system-specific nutritional and health constraints can assist with the devising of system-wide interventions and strategies. This is especially the case in the current era where increased emphasis is placed on making agriculture more nutrition- and health-sensitive. Linkages between farming systems and health and nutritional outcomes are expected through two main channels – the supply of nutritious foods and the spread of diseases and infections. Preliminary results show significant heterogeneity in health and nutrition across Africa’s diverse farming system, which may merit a system-specific approach for ensuring food and nutritional security in the region.
The Role of Livestock Ownership in Child Nutrition in East Africa
Catherine Pfeifer, Research Center for Organic Agriculture
Introduction
The objective of this paper is to investigate the influence of livestock ownership on child dietary outcomes in view of understanding how policies in the livestock sector might contribute to reduced malnutrition. Two pathways are investigated. First, the direct link between improved nutrition and livestock ownership is investigated by linking cattle ownership with consumption of milk and chicken ownership with egg consumption. Second, the indirect effect of livestock ownership and nutrition through improved income is explored by linking the nutritional diversity score of children with livestock ownership against those without.
Methods
This paper makes use of the most recent Demographic Health Surveys (DHS) in East Africa that contain a 24-hour recall of child nutrition as well as information about livestock ownership at household level. In addition, for each DHS observation, geographical layers are queried to extract location-specific information about bio-physical and socioeconomic contexts, namely the climatic agro-ecological conditions and market access. Econometric models are fitted to identify the drivers that explain milk and egg consumption and diversity score, including livestock ownership.
Findings and Interpretations
Preliminary results show that children in cattle-owning household have a much higher probability to drink milk. Also, in dairy high-potential areas such as the East African highlands, children are much more likely to get milk in general, while at the same time proximity to market decreases a child’s probability to get milk. Likewise, children living in chicken-owning or wealthier households are more likely to eat eggs. Unlike drinking milk, proximity to market enhances the chance of a child to eat eggs.
Finally, DHS data suggests that diet diversity increases with market access, wealth, and number of livestock owned.
Conclusions
Increasing evidence shows that animal-sourced food plays a critical role in reducing child malnutrition in low-income countries. Simultaneously, food systems in these countries are rapidly changing as a result of urbanization and increased wealth. Livestock production systems at farm level respond to those changes often by enhancing livestock productivity.
These transformations offer an unprecedented opportunity to improve the income of livestock keepers as well as to improve the availability of animal-sourced food and therefore could address the issue of child malnutrition in the region. The preliminary results show that both better incomes and livestock ownership are likely to improve children’s diets. However, improved market access can influence children's nutrition in two opposite ways: improved access to different food items can lead to a more diversified nutrition, but it also reduces children’s chances to get milk, as this high-value product that used to be for home consumption gets sold in the market.
Adoption of a Government-Led Homestead Gardening Program in Ethiopia
Kalle Hirvonen, IFPRI
Introduction
Since their development by Helen Keller International in Bangladesh in the 1980s, homestead gardening (HG) programs have become increasingly popular throughout the developing world. However, HG evaluations have solely focused on NGO-led interventions, many of which were not implemented at scale or across diverse agroecologies or market access settings. In this paper we explore the implementation of a government-led nationwide HG program (HGP) in Ethiopia that involved implementation by health extension workers. Our objectives are (1) to study barriers to adoption of HGPs, and (2) to assess the extent to which HGPs can improve dietary diversity in different environments.
Methods
This study is based on primary survey data collected in 2017 on households with young children (under 24 months of age). The survey is large – more than 2,500 households – and widespread, covering 88 districts (woredas) and 264 sub-districts (kebeles) in four regions of the country. In addition, the survey team conducted in-depth interviews with 264 front-line health workers working in the same localities. We use descriptive statistics and regression analysis to understand patterns of HGP adoption among poor rural households in Ethiopia, the extent of HG promotion by front-line health workers, and explore the potential impact of HGPs on dietary diversity after controlling for confounding factors.
Findings and Interpretations
More than 85 percent of front-line health workers reported promoting HGs to their clients in the past 12 months. Despite this, only 25 percent of households reported cultivating a homestead garden during the same period. Both households and health workers quoted poor access to water as one of the most important reasons for low adoption rates, and households located in areas characterized by ample rainfall and short distance to a water point are considerably more likely to adopt HGs. Households with better access to food markets are also more likely to adopt homestead gardens, as are wealthier and larger households. Results suggest that children from households that operate an HG have marginally higher dietary diversity and are marginally more likely to consume green leafy vegetables. We also find that poor access to water and poor access to markets are relatively highly correlated.
Conclusions
Both poor access (physical/financial) to food and limited knowledge of the nutritional importance of nutrient-rich foods are key constraints to improving dietary diversity in poor countries (GNR, 2017). The promotion of HGs can, in principle, address supply-side constraints. However, this policy prescription assumes that: (1) food markets are highly imperfect such that households need to produce additional foods themselves; and (2) that households have access to all necessary inputs (water, land and ideal agroecological conditions) needed to produce nutritious foods.
Our results show that households with poor access to food markets and poor access to water are less likely to operate a HG. Moreover, poor access to food markets and poor access to water are highly correlated. Together, these findings suggest that the more remote households who would theoretically benefit most from HG production are not able to do so because of limited access to water. This implies that HGs may not be a 'low-hanging fruit' to address poor dietary diversity in rural Ethiopia; improving water access is an important but potentially costly prerequisite for the adoption and effectiveness of HG production.
Effect of Home Gardening on Food Security and Diet Quality of Households in Ijebu-Igbo, Nigeria
Oluwaseun Ariyo, University of Ibadan
Introduction
Sustainable access to adequate, safe, and nutritious food is required for a healthy and active life, either through own production or purchase. However, food poverty and poor diet quality remain high in many developing countries with an increasing burden of diet-related non-communicable diseases. Ensuring food security and a high-quality diet is therefore essential to promote health in low-income settings. Home gardening’s potential to enhance food security is known but impacts on diet quality remain unclear. This study was designed to assess the effects of home gardening on food security and diet quality of households in Ijebu-Igbo, Nigeria.
Methods
This comparative cross-sectional study involved 387 women of reproductive age selected using a cluster sampling of the four local council development areas in Ijebu-Igbo Community, Ogun State, Nigeria. Based on home gardening practice, respondents were stratified as Non-Home Gardeners (NHG) and Home Gardeners (HG). Information was collected on socio-demographic characteristics, food security, and 24-hour dietary recall using a semi-structured, interviewer-administered questionnaire. Food security was assessed using the Household Food Insecurity Access Scale classified as food secure, mild, moderately, and severely food insecure; and dichotomized into food secure and food insecure for regression analysis. A multiple-pass 24-hour diet recall was conducted for all respondents for two weekdays and one weekend. The 24-hour diet recall data were used to calculate diet quality using the Diet Quality Index International method categorized as high (>60), medium (41-60), and low (0-40). Body weight and height were assessed to determine body mass index and categorised using World Health Organization standards. Data were analyzed using descriptive statistics, Chi-square test, and binary logistic regression analysis at p=0.05. Four focus group discussions of 10 respondents each were held with Home Gardeners to identify likely pathways of influence of HG on FS and DQ and analyzed thematically.
Findings and Interpretations
Respondents’ age (years) was 40.4±12.8 and 41.7±15.9 for NHG and HG. Majority in both NHG (72.1%) and HG (73.9%) were traders, and incomes between ₦10000-₦30000 were most common in both NHG (54.8%) and HG (51.3%); groups were not statistically different. Only 29.2 percent of respondents practiced HG, duration of practice was 6.96±5.57years; and commonly grown crops were staples (44.0%), vegetables (26.0%), and medicinal plants. Food security was significantly higher among HG (31.3%) compared to NHG (15.1%) and severe food insecurity was four times higher among NHG (33.1%) compared to HG (8.7%). High diet quality was higher among HG (58.8%) than NHG (53.9%). Overweight (33.1%; 32.0%) and obesity (23.5%; 21.0%) were higher among NHG than HG and underweight (5.9%) occurred exclusively among NHG. Home gardeners had three times higher odds of being food secure than NHG (OR=2.24, CI= 1.07-4.69) and two times higher odds of having high quality diet (OR=1.221, CI=1.18-1.89). Focus group discussion respondents identified the benefits of home gardening to include enhanced income, increased physical access to foods, and inclusion of available food stuff during food preparation. Identified challenges to home gardening included inadequate space, limited time, and free-range domestic animals’ intrusion into the home gardens.
Conclusions
The practice of home gardening is only at 29.2 percent among households in Ijebu Igbo community, Nigeria. Home gardening practice enhances household food security and diet quality and is significantly associated with good anthropometric characteristics among women of reproductive age. Identified pathways of improvement were increased income, increased physical access to foods, and increased utilization of available foodstuffs during food preparation. Increased efforts to promote home gardening and supports including incentives and training for cultivation of nutrient dense crops are hereby recommended as part of the ongoing social protection activities in the country. Policy and strategies to ensure provision of space for home gardening activities and prohibit free range animals should be put in place to encourage home gardening.
A Participatory Agroecological Intervention Amongst Rural Tanzanian Farmers Increases Crop Diversity, Food Security, and Maternal Dietary Diversity: A Midline Evaluation
Marianne V. Santoso, Cornell University
Introduction
Our formative research in Singida, Tanzania, indicated that smallholder farmers faced several barriers to optimal nutrition: limited agricultural knowledge, high levels of food insecurity, sub-optimal infant and young child feeding practices, and gender inequity. The Singida Nutrition and Agroecology Project (SNAP-TZ; NCT02761876) is a randomized effectiveness trial of a participatory intervention to improve these outcomes. Rural farmers with children under one year of age were recruited from 20 villages (n=587). We report on the impact of SNAP-TZ on crop diversity, household food security, dietary diversity, and women’s empowerment after one year of the three-year study, as well as lessons learned from implementation.
Methods
The intervention began by selection of ‘mentor farmers’, one man and one woman from within each of the ten intervention villages (n=20). In 2016, they visited a similar project in Malawi, followed by a two-week intensive participatory course which integrated agroecology, nutrition, and gender issues. All participating farmers in intervention villages (n=279) received ongoing support from mentor farmers on these issues and legume seeds to plant in December 2016. Participants are surveyed (1) on socio-demographic, agricultural, and health annually and (2) on food security, diet, and experiences with the intervention twice annually. Qualitative data collection included participant observation of village meetings and household visits, semi-structured interviews with mentor farmers (n=20), and 10 focus group discussion with participating households (n=33).
Difference-in-difference (DID) analysis between baseline (BL) in the growing season (BLgrowing) and follow-up (FUgrowing) was used to evaluate impact on crop diversity (range:0-7) and the Abbreviated Women’s Empowerment in Agriculture Index (A-WEAI, range: 0-1). DID analysis between BLharvest and FUharvest was used to evaluate impact on the Household Food Insecurity Access Scale (HFIAS, range: 0-27), UNICEF’s Child’s Dietary Diversity Score (CDDS, range: 0-7), and Minimum Dietary Diversity for Women (MDD-W, dichotomous) in STATA14. Clustered standard error was used for all analysis.
Findings and Interpretations
In January and July of 2016 (at baseline), crop diversity, HFIAS, DDS, MDD-W and A-WEAI scores did not differ between intervention and control households, indicating effective randomization.
By January and July 2017, 88.9 percent and 64.4 percent of female and male participants, respectively, had attended mentor farmer meetings, reporting 2.7±1.6 and 2.4±1.6 meetings a month, respectively. DID analysis indicated that households in intervention villages had an increase in crop diversity (β=1.06, p<0.001), decrease in HFIAS (β=-2.08, p=0.016), and were more likely to meet MDD-W than those in control villages (OR: 2.06, p<0.001). There were no significant differences in CDDS (β=-0.04, p=0.86) or A-WEAI (β=0.01, p=0.75).
Qualitative data analysis revealed several themes: (1) mentor farmers saw the intervention as feasible based on the observed successes of their Malawian counterparts; (2) having both a male and female mentor farmer in each village is important to show that nutrition and agriculture are the concern of both men and women; (3) there is a widespread perception that the study is solely for women due to its messaging around nutrition; and (4) mentor farmers tended to put less emphasis on gendered topics, e.g. family planning, during meetings to reduce resistance from participants.
Conclusions
We have begun to observe shifts in three important indicators along the pathway to improved nutritional outcomes among women and children: crop diversification, maternal dietary diversity, and food security. We have yet to see impacts on child nutrition or gender equity. Greater emphasis on empowering both men and women participants to change these social and nutrition-related behaviors may improve these outcomes. As such, the implementing team has added renewed focus on addressing gender inequity in the program with a special gender training and continued efforts to involve men in discussion regarding nutrition and gender. Because agriculture projects are seen as men’s domain and nutrition projects as women’s domain, having both male and female mentor farmers discussing agriculture, nutrition, and social equity has been very important. Rapid qualitative and quantitative analysis to improve program implementation has also been helpful in ensuring improved quality of implementation.
Agriculture to Nutrition: Promoting Vegetable Production for Improving Dietary Diversity in Ethiopia and Tanzania
Farai Gwelo, FANRPAN
Introduction
Despite significant progress towards improving nutrition, the global community is still grappling with multiple burdens of malnutrition. Two billion people lack key micronutrients like iron and vitamin A. Evidence shows that actions delivered through the 'nutrition sector' alone can only go so far, hence, the call to address the root cause of poor nutrition. Agriculture is the backbone of diet and nutrition, therefore agricultural interventions also need to address nutrition. The Agriculture to Nutrition project, therefore, seeks to provide evidence for agriculture's potential to deliver positive nutrition outcomes in Ethiopia and Tanzania, where selected Nutrition-Sensitive Interventions (NSI) are being tested.
Methods
Agriculture to Nutrition (ATONU) is testing three pathways to deliver improved nutrition to smallholder farming households: (i) agricultural production for own consumption, (ii) use of agriculture income to purchase other nutritious foods, and (iii) women’s empowerment to improve agency and nutrition. The pilot project uses a cluster-randomized design to estimate the effect of the interventions on the primary outcome of dietary diversity among women of reproductive age in rural Ethiopia and Tanzania. However, this paper will focus on the production for own consumption pathway.
A total of 1600 farming households (800 in Ethiopia and 800 in Tanzania) with women of child-bearing age and children under the age of five were selected from 20 villages in each country. ATONU introduced a package of NSIs that included vegetable production for own consumption in order to improve the dietary diversity of participating farming households. The households were visited regularly by trained field assistants, during the evaluation period from April to December 2017 for delivery and re-enforcement of ATONU messages and collection of process data. Both quantitative and qualitative data were collected on agricultural production, food consumption patterns, and vegetable production and consumption at the baseline and after one year of interventions.
Findings and Interpretations
The baseline survey revealed that the proportion of households that were involved in vegetable production was less than 10 percent in Ethiopia and less than 40 percent in Tanzania. However, following a year of delivering nutrition-sensitive interventions, the uptake of vegetable production increased to ~84 percent in Tanzania and 100 percent in Ethiopia. The results also showed increased consumption of vegetables (especially the dark green vegetables) from 35 percent to 93 percent in Tanzania and from 20 percent to 100 percent in Ethiopia. Vegetable production contributed to increased vegetable consumption as 71 percent of the households consumed vegetables from their own gardens, whilst only 29 percent of the households sourced their vegetables from the market. Although some households produced up to six different types of vegetables, most households produced an average of two different vegetables in Tanzania, whilst in Ethiopia, each household produced an average of three different types of vegetables following the introduction of NSIs. Green leafy vegetable varieties were the most produced in both countries.
Conclusions
The focus of this intervention was to make sure that households produced some vegetables for home consumption in order to improve their dietary diversity rather than for sale. The uptake of this intervention was a success, and most households (96 percent of producers) produced enough vegetables for their own consumption, with a surplus for sale. For those households that did not produce any vegetables, the main reasons cited were lack of water, lack of land, lack of time, and chickens and wild birds eating the vegetables; hence those households preferred to buy rather than produce. Households are being encouraged to use some of the income from vegetable sales to buy more seed so as to make the practice more sustainable. It is hoped that the increase in vegetable consumption will contribute significantly to dietary diversity of women of child-bearing age and children under the age of five.
