Session 8A: Food environments and drivers of food choice
byANH Academy
Academy Week Research Conference
| Agriculture, Nutrition, Public Health
Date and Time
From: 28 June 2019, 14:30
To: 28 June 2019, 15:50
BST British Summer Time GMT+1:00
Country: India
Open Full Event


Five, 10-minute abstract-driven presentations.  

Speakers and Presentations:


  • Chair: Christine Blake, University of South Carolina
  • Alysa Grude, ACDI/VOCA
    Increasing the productivity and market linkages among smallholder dairy producers in Bangladesh: The effects on total amount sold versus set aside for home consumption
    Slides/ Recording

  • Reshma Roshania, Emory University
    Food environments and child nutrition among circular migrant families working in the brick kilns of Bihar, India
    Slides/ Recording

  • Christopher Turner, London School of Hygiene and Tropical Medicine
    Transitioning food environments, transitioning livelihoods: A qualitative investigation of food acquisition practices from two rapidly urbanising villages in peri-urban Hyderabad, India
    Slides/ Recording

  • Dare Akerele, Federal University of Agriculture Abeokuta
    Food Away from Home in Nigeria: Consumption, Drivers and Nutritional Implications of Within-Day Meals

  • Gowthami Venkateswaran, University of Illinois
    The Domino’s Effect: Impact of Fast Food Chains on Urban Health in India
    Slides/ Recording



Increasing the productivity and market linkages among smallholder dairy producers in Bangladesh: The effects on total amount sold versus set aside for home consumption

Alysa Grude, ACDI/VOCA 

Introduction:  Nutrition-sensitive agricultural projects often assume that as productivity of nutrient-rich foods increase that household consumption of that particular food subsequently increases. However, these approaches often fail to consider the income-consumption trade-offs and the different market linkages that support production for sale as the project increases market opportunities. The USAID funded and ACDI/VOCA implemented Livestock Production for Improved Nutrition (LPIN) project aims to increase smallholder farmers’ dairy productivity for increased incomes and consumption of milk and dairy products in Bangladesh for improved nutrition. Project specialists sought to understand the relationship of milk production and milk consumption among project beneficiaries, presented here.

Methods:  The sample analyzed in this study included 322 smallholder livestock rearing households in the zone of influence of the LPIN project in Bangladesh. Data collection occurred for the purpose of the annual survey and report. Following cleaning and coding of the data, binary and multivariate linear and logistic regression analyses were conducted in R Studio to examine the cross-sectional association between total milk production and the explanatory variables (e.g., percent consumed, total milk consumed).

Findings:  As total milk production increases, the percent of milk that is set aside for home consumption decreases. Data analysis showed that milk production under 1 liter a day was significantly associated with an increased likelihood that 50% or more of the milk was set aside for home consumption. This suggests that low production of milk is largely set aside for home consumption. This may be due to the low volume of milk not being attractive to milk collectors. However, as total milk production increases, the percentage of milk set aside (opposed to sale or gifting) for home consumption decreases (increase in one liter was associated with a 13.5% reduction in total percentage set aside for home consumption: (p- value 6.57e-15 ***). This may suggest that as milk production increases, farmers have the “milk capital” to engage with milk collectors for sale of their milk. Further, only 5.3% of households were setting aside an average of 1 liter or more a day, which demonstrates that despite an overall increase in milk production, there remains a need to increase consumption.

Conclusions:  From a project reflection standpoint, this analysis suggests that milk consumption does not increase as much as was initially expected just by increasing farmers’ milk production. Nutrition-sensitive agricultural projects aiming to increase household access to and consumption of nutrient-rich foods should be aware that market system improvements, such as increased production and linkages to aggregators, may foster preference towards sale of production, rather than preference towards consumption.


Food environments and child nutrition among circular migrant families working in the brick kilns of Bihar, India

Reshma Roshania, Emory University

Introduction:  Bihar contributes substantially to India’s overall burden of undernutrition. Short-term circular migration is an important livelihood diversification strategy for poor marginalized households, often in addition to rain-fed subsistence agriculture. The brick industry in Bihar operates seasonally and predominantly on migrant family labour. Young children who engage in circular migration are vulnerable to poor health and nutrition given exploitation, loss of food security benefits, isolation from health services, and hazardous living conditions. We explored child nutrition status and the food environment, including access to food security benefits, among circular migrants, focusing on differences between interstate and intrastate origin.

Methods:  We conducted this mixed-methods study in June 2018. The qualitative component consisted of 14 in-depth interviews with circular migrant parents and 8 key informant interviews with kiln managers and labour contractors, conducted in 8 kilns across 3 districts, purposively selected for regional variation. The interviews explored perceived changes between home and destination regarding diet and food environment: availability, affordability and accessibility. We performed code- and case-based thematic analyses in MAXQDA.The quantitative component employed a stratified, cluster sampling design and consisted of a cross-sectional survey in 552 randomly selected brick kilns (clusters) throughout Bihar. The sampling frame was obtained from district Departments of Mines and Geology. We digitally collected kiln, household and child-level data. Per kiln, three circular migrant households with children 0-35 months of age were randomly selected. Anthropometric measures were collected on each sampled child (n = 1198). Food security was measured using the Food Insecurity Experience Scale. Descriptive, bivariate and logistic regression analyses were conducted in SAS. The primary exposure of interest was intrastate vs. interstate origin; the primary outcomes were stunting (< -2 SD height-for-age z score) and wasting (< -2 SD weight-for-height z score); covariates included wealth index, child age, gender and parity.

Findings:  Qualitative findings indicated that many migrants are subsistence farmers, additionally engaging in circular labour migration for reasons including insufficient income, debt, and lack of irrigation infrastructure. Many migrants spoke of higher food prices at their destination; moreover, unlike at their home where food was obtained from own cultivation and Public Distribution System rations, the private market was their only source of food at the brick kilns. Perceived food affordability increased during migration, likely because of regular monetary payment mechanisms at kilns. Migrants did not perceive any differences in food availability between their home and destination. Differences by origin in perceived food environment mainly related to desirability and food preferences. Quantitative results showed food security was positively associated with wealth and land ownership, and negatively associated with PDS utilization. Those who had migrated for more years were less wealthy and less food secure. Among children 6-23 months, 13% had a minimum acceptable diet; MAD was higher among interstate migrants (17%), compared to intrastate migrants (10%) (p=0.14). Prevalence of stunting was lower among interstate migrants (47%) compared to intrastate migrants (55%, aOR: 0.66, 95%CI: 0.50-0.88) In contrast, wasting was higher among interstate migrants (43%) compared to intrastate migrants (34%, aOR:1.51, 95%CI: 1.17-1.94).

Conclusions:  Circular migration is a growing response to socio-economic structures that perpetuate inequity and undernutrition in India. Our results contribute to the ongoing policy discourse on the portability of PDS, suggesting that engagement in circular migration does not necessarily lead to accumulation of wealth and long-term food security; policies to protect migrants’ right to food security benefits should be implemented. State of origin emerged as an important predictor of nutrition status, operating differently for acute and chronic malnutrition; we will further explore pathways of nutrition by origin in future analyses.


Transitioning food environments, transitioning livelihoods: A qualitative investigation of food acquisition practices from two rapidly urbanising villages in peri-urban Hyderabad, India

Christopher Turner, London School of Hygiene and Tropical Medicine

Introduction:  Food environment research is increasingly gaining prominence in low- and middle-income countries (LMICs). Policymakers seeking to tackle food and nutrition security and the double burden of malnutrition are turning attention to the role that food environments play in shaping diets, nutrition and health. This qualitative study reveals in-depth perceptions and experiences of food environments and food acquisition practices from two rapidly urbanising villages in peri-urban Hyderabad, India. It is also the first study to apply, validate, and sensitise concepts from the ANH Academy’s food environment conceptual framework. Methods include in-depth interviews and an innovative qualitative geographical information systems (QGIS) approach.

Methods:  Two villages in peri-urban Hyderabad, Patelguda and Thumaloor, were purposively selected to provide a sampling frame of households in a rapidly urbanising setting. Households with at least one adult male and female aged 18-65 were eligible for inclusion. Forty households were randomly sampled and allocated to one of two qualitative methods: 1) in-depth interviews (n=18); 2) an innovative Q-GIS approach featuring three-day participatory photo mapping (PPM) and follow-up photo-elicitation interviews (PEI) (n=22). Isolated interviews were conducted with one male and female participant from each household, guided by the overarching question “If I were to live with you, what would we eat, where would we get foods from, what would we see, what would we do, and who would we do it with?”. Topic guides focused on perceptions and experiences of the food environment, interactions between personal and external domains and dimensions, and intra-household gendered food acquisition practices. Analysis featured deductive and inductive techniques in an iterative process. Elements of grounded theory and thematic analysis generated key themes. Comparison between these themes and the ANH Academy food environment conceptual framework sensitised concepts, established conceptual validity, and identified existing gaps.

Findings:  Overall, findings reveal the ways in which people navigate personal and external food environment domains and dimensions to acquire food from diverse sources as part of daily life in transitioning LMIC settings. Participants revealed how transitioning food environments and livelihood strategies are entwined with shifting food acquisition practices within this rapidly urbanising context. Key perceived drivers of change in recent years included the sale of agricultural land for urban development, considered to have increased the local availability and reliance upon market-based food vendors, and also the rise in ownership of personal motorised vehicles that has both improved accessibility to distant markets and created the ability to transport larger quantities of food. Participants also evoked a sense of community and trust, referring to local vendors as ‘known people’. These sentiments of trust were closely connected with the practice of acquiring food on credit, as well as food safety discourse connected to concerns around food adulteration and pesticide use, highlighting the importance of social contracts in food acquisition practices. Narratives of intra-household gendered food acquisition practices were also found to be prevalent, and included differential activity space patterns related to work-based travel and the influence of children and peers on food acquisition.

Conclusions:  This study contributes contextualised knowledge and understanding of food acquisition and consumption behaviour from two rapidly urbanising villages in peri-urban Hyderabad. The triangulation of participant’s narratives, photographs and maps represent emic perspectives and experiences of interactions between external and personal food environment domains. Findings have direct implications for public health nutrition across transitioning LMIC settings, as well as wider relevance through the application and sensitization of the ANH Academy’s globally applicable food environment framework. Themes including ‘sense of community and trust’ and ‘intra-household gendered food acquisition’ provide new critical insights for food environment research in LMICs.


Food Away from Home in Nigeria: Consumption, Drivers and Nutritional Implications of Within-Day Meals

Dare Akerele, Federal University of Agriculture Abeokuta

Introduction:  Food away from home (FAFH) have been progressively playing crucial role in the diets of many households in developing countries, Nigeria inclusive. A number of these foods have been reported to be low in micronutrients, high in fat/cholesterol, and energy dense, with consumption attributed to higher prevalence of malnutrition and diet-related non-communicable diseases. Determinants of purchase decisions and consumption of FAFH are less empirically studied in Nigeria, and little is known about what the specificity of consuming FAFH and within-day meals (breakfast, lunch, dinner) hold for nutrient availability. Findings are important for public health nutrition programming.

Methods:  We used the Nigeria Living Standard Measurement Survey nationally representative household panel data for 2015/16 comprising two time periods (six month intervals) for analysis. Relevant data used for analysis include, among others, household socio-demographic characteristics, income, weekly value of aggregate foods consumed AFH, values of specific food groups (snacks, dairy-based beverages and vegetables) consumed AFH, expenditure on foods and non-food items, and quantities of 120 different food items consumed from which calories and selected nutrients were estimated. Whether (or not) households consumed breakfast, lunch, dinner, or side-dishes AFH were captured. Descriptive statistics were used to summarize data, while bootstrap double-hurdle regression (that addresses possible unobserved heterogeneity due to time dynamics in food consumption) was employed to identify key factors determining decisions to consume food AFH and the share of household budget expended on the foods. The estimated regression (econometrics) model is stated as:

H_it = ∑βX_kit  + α_i +e_it

where H_it captures the budget share spent on either aggregate food, or specific food groups (snacks, dairy-based beverages, vegetables) consumed AFH respectively. X_kit  is a vector of determinants such as socio-demographic characteristics, income, location/regional factors, among others;  α_(i ) represents household specific effects, while e_it is the error term.

Findings:  Results (descriptive statistics) show that on average, households spent approximately 28% of their total budget on food AFH.  For the aggregate food, and the range of food groups (vegetables, dairy-based beverages and snacks) considered, consumption AFH is higher for urban and non-agricultural households, as well as households whose heads had tertiary education. The amount expended on them also increased progressively with higher income levels. This buttresses income growth, urbanization, higher education attainment, and opportunities for non-farm jobs/occupations as vital triggers for consumption of food AFH. Regression results also suggest that the likelihood of households consuming one or more of the selected food groups, and the share of household budget expended on them respectively, will increase with higher income, hours spent on non-farm businesses, urbanization and the presence of children under five years of age and/or adolescents in the household. Consuming food AFH appears to be connected with higher consumption/availability of fats, and lower availability of iron and calcium. Whether households consumed (or did not consume) breakfast or lunch AFH seems to hold little significance for daily per capita calories, proteins and fat consumed by households. However, side dishes and dinner AFH may make a difference. Findings hold important implications for nutrition and health.

Conclusions:  Using nationally representative household panel data, we examined, among others, the determinants of consumption of FAFH, and the implications of within-day time of consumption for nutrient availability. We found that increased income, urbanization, higher education attainment, household composition, opportunities for non-farm jobs/occupations and hours spent on non-farm businesses are key determinants of purchase decisions and/or consumption of FAFH. Advancing consumption of FAFH may mean less availability of iron and calcium for households. Taking breakfast or lunch AFH seems to hold little consequence for daily calories, proteins and fat consumed by households while side-dishes and dinner AFH may trigger nontrivial divergence.


The Domino’s Effect: Impact of Fast Food Chains on Urban Health in India

Gowthami Venkateswaran, University of Illinois

Introduction:  In this paper, we look at the impact of access to unhealthy food on urban health in developing countries. To address the issue of double burden of malnutrition in developing countries, we examine potential policy options in India using a unique identification strategy. We evaluate two possible policy responses – “sin tax” on unhealthy foods and stricter zoning laws for fast food chains – to understand the effectiveness of these policies on urban obesity. Our findings suggest modest gains from sin tax policies and significantly large gains from implementing stricter zoning laws for fast food chains.

Methods:  The disease burden placed by excessive consumption of junk food is an important public health issue. Many developed and developing countries have considered imposing a sin tax on food products high in sugar and calories. In 2016, the southern state of Kerala, with the second highest obesity rate in India, became the first Indian state to impose a 14.5 percent sin tax on food products high in sugar and calories. Using data from the Demographic and Health Survey (DHS) and National Sample Survey (NSS), we use a double difference strategy to evaluate the impact of the new tax policy on urban health. We look at outcome variables like Body Mass Index (BMI), blood pressure, blood glucose and hemoglobin levels, among others. Policymakers have also considered enforcing stricter zoning laws around schools, playgrounds, residential areas, etc., to curb childhood obesity. To evaluate and compare the sin tax policy to a potential zoning law policy, we use spatial data from the online restaurant review website, Zomato, to construct a fast food restaurant-density variable for each DHS cluster. We use this restaurant-density variable as an instrument for individual households to understand the impact of vicinity to fast food chains on childhood and adult obesity.

Findings:  Our findings suggest that, while the tax on sugary drinks and calorie-rich food had modest impacts on urban obesity, proximity to fast food chains results in significantly negative health outcomes, particularly for children. These findings make a case in favor of stricter zoning laws around schools, playgrounds, etc., where direct access to unhealthy food for children might have serious negative outcomes for childhood obesity. Following the example of the state of Kerala, other states like Gujarat are considering a sin tax as a policy measure to curb obesity rates, particularly in urban areas. As the federal government evaluates newly proposed regulations for a potential national level tax on unhealthy foods, it faces growing pressure from fast food chains which bring in billions of dollars in foreign direct investment to a sector currently valued at USD 57 billion. To justify any regulatory policy measure, it is imperative to understand the cost currently imposed by consumption of unhealthy food. Our study evaluates the disease burden placed by the consumption of unhealthy food, while also evaluating potential policy solutions.

Conclusions:  Even though the number of malnourished children in India is double that of sub-Saharan Africa, it also ranks third in the world for obesity, after the U.S. and China, and second in the world for type 2 diabetes. This paper looks at the current issues related to malnutrition in urban areas, the urban food environment, the nutrition transition and the double burden of malnutrition in fast-growing economies like India, while also evaluating and proposing policy solutions.

Login or Join ANH Academy to be part of the conversation