Correction: Tracking a decade of food system transformation in Vietnam's Mekong Delta: production, consumption, and nutrition

The global food system is undergoing rapid transformations due to factors such as economic development, urbanization, climate change, and trade liberalization. These changes influence agricultural production and household food consumption, affecting dietary patterns and nutrition outcomes in complex and interconnected ways.On the supply side, agricultural production has expanded in scale and diversity, but also faces mounting constraints due to environmental degradation and climate-related shocks (Springmann et al., 2018). Expanding agricultural land uses and intensifying farming practices have boosted yields and enhanced food availability, but often at the cost of excessive input uses, water and soil depletion, biodiversity loss, and increased greenhouse gas emissions (Ramankutty et al., 2018;Tilman et al., 2011). Additionally, the rise of globalized supply chains, dominated by supermarkets and large-scale retailers, has transformed food distribution, increasing access to diverse products but often leading to the prioritization of ultraprocessed foods and marginalization of small farmers (Reardon et al., 2019). Regarding food consumption patterns, diet transitions tend to move from traditional, localized diets rich in vegetables, fruits, and legumes to a more modern, globalized diet marked by increased consumption of animal-based foods, fats, added sugars, and processed products (Popkin, 2017). The availability and consumption of ultra-processed foods (UPFs) have surged globally, now contributing 50-60% of daily energy intake in some high-income countries, with low-and middle-income countries (LMICs) following similar trends (Scrinis & Monteiro, 2022). Likewise, the intake of sugar-sweetened beverages (SSBs) has risen sharply worldwide (Vandevijvere et al., 2019). Additionally, global meat consumption increased by 20 kg per capita between 1961 and 2014 (Vermeulen et al., 2020).These shifts have significant implications for nutrition outcomes. Globally, overweight and obesity are on the increase, with around 40% of adults and 20% of children now classified as overweight or obese. The prevalence of obesity and diet-related non-communicable diseases (NCDs) is also rising (UK Development Initiatives, 2022). In LMICs, higher socioeconomic status (SES) and urban residency are linked to healthier dietary patterns but also greater energy, cholesterol, and saturated fat intake (Mayén et al., 2014).While substantial research has explored food systems, most food studies focus on examining one or two aspects -supply, demand, or nutrition outcomes, but rarely analyzed comprehensively their dynamic interconnections (Brouwer et al., 2020). This fragmented approach limits our understanding of how supply, demand, and nutrition outcomes interact over time to shape dietary patterns and health impacts, highlighting the need for more longitudinal studies to integrate all three components (Even et al., 2024). This study contributes to addressing that gap by examining how changes in supply-side and demand-side have influenced household diets and nutrition outcomes over the decade in a LMIC context.Our focus is on the Mekong River Delta (MRD) in Vietnam, a highly fertile and biodiverse food-producing region. However, millions of small-scale producers face mounting risks from climate change, exacerbated by unsustainable development (Edmonds et al., 2020). Effects from climate change could cause more households in the MRD to not be able to grow food and increase risk of undernutrition in kids, poor and landless households (Headey & Ruel, 2023). To better understand diet transformation and nutrient gaps in Vietnam, particularly in the MRD, amid climate change, urbanization, and migration pressures, this study seeks to address two main research questions: (Q1) How did the pattern of food production and consumption of households in the MRD of Vietnam changed over the years? and (Q2) How has these changes shaped household dietary and nutrition shifts? By exploring these questions, we use multiple national datasets to examine how both supply-side and demand-side changes have shaped household's diet composition and nutrition outcomes over time.The remainder of this paper is organized as follows: first, we describe the methods and data sources, followed by results examining production patterns, consumption changes, and nutrition outcomes, and conclude with a discussion of the integrated findings and their implications for sustainable food system transformation.In this analysis, we use multiple datasets. The main dataset used is the Vietnam Household Living Standards Survey (VHLSS), collected every two years since 1992 by Vietnam General Statistics Office (GSO). The VHLSS data covers various aspects, including household and respondent characteristics, food production, and food consumption as well as expenditure. We specifically focus on data from 2010 to 2020, as it provides detailed measures of food consumption over the past month. Additionally, we examine both household income and expenditure to gain a comprehensive view of household production, income, and food consumption. This repeated cross-sectional dataset comprises nearly 9,400 households nationwide and 1,760 households in the MRD during each round. The survey sample is designed to represent the entire country, its regions, and provinces. It includes 9,000 survey clusters selected from the area sampling frame of the Population and Housing Census. These clusters are drawn from 3,000 communes or wards, with three clusters chosen from each commune or ward. The sampling procedure of the VHLSS follows three stages:Step 1. Stratification: Provinces and cities are classified into six socio-economic regions, namely the Red River Delta (RD), Northern Midlands and Mountains (NM), North Central Coast and Central Coast (NC), Central Highlands (CH), Southeast (SE), and Mekong River Delta (MRD). This stratification ensures a representative sample size for each region.Step 2. Selection of survey clusters (Primary Sampling Units): Villages, hamlets, or residential groups are systematically and randomly selected from the area sampling frame of the Population and Housing Census.Step 3. Household selection (Secondary Sampling Units): In each selected cluster, approximately 20-25 households are chosen to participate in the survey through simple random sampling based on the household listing within the cluster.To monitor trends over time while minimizing respondent burden, the VHLSS employs a panel rotation technique. Specifically, half of the households in each cluster are retained for the next survey round, while the remaining half are replaced with newly selected households.In addition, we analyze provincial-and regional-level food production data from the Vietnam Statistical Yearbook (SYB) from 2010 to 2020. This annual statistical survey provides data on the socio-economic situation at the provincial, regional, and national levels, and is implemented by the provincial and national statistical offices.Finally, we utilize the General Nutrition Survey (GNS) to explore food consumption, nutrient adequacy, and nutrition status in 2010 and 2020. The survey, jointly implemented every ten years by the National Institute of Nutrition (NIN) and the GSO, includes around 22,000 households from 25 provinces and cities, representing six ecological regions across the country.From the VHLSS data, each household reported the quantity of food items that they produced the previous 12 months before the survey. The quantity of food produced by each household was then calculated in kilograms (kg). For most food items, this was divided into raw and semi-processed forms. For a few items measured in different units, conversion to kg was applied. For example, one liter of milk equals 1.03 kg (USDA, 1992), and 28 eggs (poultry) equals one kg (Dang et al., 2022).For food production data, our analysis includes only five food groups: staples, vegetables, fruits, animalsourced food (ASF) such as meat, dairy, and eggs, and aquatic products such as fish and seafood. Meanwhile, food consumption data also includes additional categories such as sweets and beverages (which includes sugar, confectionery, alcohol, and packaged drinks), oils and spices. Due to the high variability in the quantity of food produced by households, we use data from the VHLSS only to calculate the average proportion of food that is produced and sold by households in the production section.To explore the trends on agriculture production quantity, area and productivity in Vietnam, we utilized the SYB published by GSO for the same period as the VHLSS. These data have information on production quantity, area and productivity, which are representative at sub-national and national levels. These data help inform if the changes over time are due to increasing produced food amount, expanding agricultural land or improved productivity.The monthly amount of food consumed was calculated a similar units, and as for food For consumption, to the of one liter kg (USDA, In to our results to we the consumption to the average consumption per This is the conversion that each household and & et al., production and consumption are also measured in in to In the VHLSS food consumption is a for This includes food and beverages and consumed the by household they in the household or are or the survey not the quantity of it only the expenditure on these To examine expenditure we in household food expenditure from 2010 to in across food groups: staples, vegetables, fruits, fish and and and and We calculated the of each food expenditure to the food expenditure as of household patterns, this as a & 2018). To this data, we the expenditure the conversion as a for 2024). Due to the in the VHLSS data, which not the quantity of food consumed the we used nutrient intake data from the to a more comprehensive understanding of dietary in the calculated based on the of food and food groups & is a used and to & as it the of food or food groups produced within a such as a in this a or a a For production by food we the same food groups used in the Household & In this analysis, we calculate based on nearly and aquatic that farmers household dietary diversity, we the dietary biodiversity that the of food intake by the of and consumed by a household & et al., 2018). Additionally, household dietary was measured the which as a for a household's economic to access a of foods and is of its food The is based on 12 food groups: and and and and and oils and & energy and nutrient calculated for energy, and by the or national average daily intake by the daily intake per et al., The for energy, for for for for and for et al., The gap was then by the from 1. These the to which the average intake or of the for each nutrient and outcomes are a to but these also not in the VHLSS and For that we a the data from 2010 and 2020. The a national sample size of in 2010 and of in 2020. We the status of children five and years and and by status of children five years is and with a two are classified as which with a two are classified as with a two are Meanwhile, with are overweight or et al., children status is and with a two are with a two are classified as with a one are overweight or on the of the et al., adults to status is with a to or greater kilograms per are classified as overweight or based on the for on VHLSS data was by for and and or in the This paper that are as or and the changes in food production, consumption, and nutrition in the MRD over the past decade It is to that survey in all to the results are representative of the analyzed was to examine between dietary and food system is for data of & 2011). was the in with the to and selection was by the et al., for including household income, of the household and migration The was chosen and for the 2020). with for and dietary outcomes. The to the results for six different regions as well as the national the MRD with regions and the national average the of this as well as in agricultural production food consumption, and These findings inform the of agricultural and the development of that food and sustainable with some of households in different regions of Vietnam and the MRD over we explore food production, food consumption, and 2010 and 2020, as in the MRD the of households households average income and the proportion of households with six regions of It was the with 40% of household or as high as in and only of household or higher (which is approximately half to of the in the changes of household over time in the The proportion of households in and in food production by and due to urbanization, climate change, and The proportion of households has increased by which be linked to migration patterns, economic or changes in household household improved over with the of or a by The average household size at while income per capita and the of households by in the same time to the has by which could be to or expanding of climate change, between and MRD rising and have risen by average of per in increased and more It is that of the Delta could be at risk of by 2020). These shifts including shifts in agricultural economic and on migration patterns, and health in over the past farmers in the MRD have produced the most staples, vegetables, fruits, and aquatic products to The produced to six as as the that was increase in the production over time in the MRD in all food groups production increased from 2010 to 2014 and then The food groups increased in production per household each in agricultural production across all regions focus on The MRD as the with the production area across all food particularly for and in the past the agricultural production area in the MRD has expanded across all food as in The in production area was for followed by and aquatic while a increase and this has the production, the productivity of staples, and aquatic products in the MRD are the national average in In this the for in productivity across all food groups have increased over with the most in and aquatic products per in with the the MRD has the proportion of sold food across most food as in the proportion of food sold has increased by for staples, for fruits, and for foods Meanwhile, the for aquatic products and have over the past of production we measured it by both food groups and food The MRD as the diverse following the which is by households in the MRD only food groups and food In the Northern Midlands and the production diversity, with nearly food groups and food per a in production is across all regions of the country, both in of food and food groups for examine factors with by food MRD households, was that every increase in the household and each additional household with and higher by food households higher and households higher and while households with produced more food in household income, and also influenced by food increase of in household income by food in production household and size increased by and households produced more food groups the a produced higher food to households the and households all higher food in of household expenditure across various categories is in households in the MRD a proportion of their expenditure to aquatic foods and sweets and the national Meanwhile, the in food expenditure over time in the MRD be in The households in this have their expenditure on 12 and increased their expenditure on and and six from 2010 to 2020. the food groups in which they expenditure in the and did not change over time in of expenditure. These trends could be due to a diet transformation and households their but they could also be due to changes in from the 2010 and 2020, which includes both and consumption in that MRD have increased their intake of and since However, the average intake of and the of per per trends increase in the consumption of foods, which now the national and a rise in consumption. Additionally, the consumption of sweets in this has increased in 2020, which is the national average to all households in the survey, the proportion of food consumed by households that was across all food groups and over the years the consumption from vegetables, and is the average daily nutrient in 2010 and for adults are The how the average consumption to the of each or In of energy, the average consumption was 20% the in with a to in The Central Highlands was the only energy consumption over In consumption of both and the and both increased in 2020. However, average consumption, which was the in to the in household dietary diversity, the results that the MRD the dietary in the country, following the Northern Midlands and with and of and In the Southeast the dietary diversity, with and average dietary improved over the past with increasing from to and from to However, in the MRD, both at approximately for and for for from the that increase of in annual household income is with higher one additional food in consumption and each additional household the by or higher higher to with households urban results for in that a rise in annual household income consumption by one food and each additional household by the a or higher higher with In households that is farming or in their National Institute of Nutrition (NIN) also measured outcomes over the same time period that we study in this the in the of children years of that are and regions the Northern Midlands and a in between 2010 and 2020, with the MRD the of regions a in the of children five years MRD the in the prevalence is at as classified by et al., the in the of and are overweight between 2010 and 2020. regions increase with the of are overweight increasing more the of are The Mekong River Delta the increase for both and at and for and data, this study a comprehensive of food system changes in the MRD over the past The analysis on three production, consumption, and nutrition patterns, while also for changes in the gaps and these provides for to a healthier and more sustainable transformation of the food system in the results that the MRD has in all of children However, particularly in as well as rising of overweight and obesity and health These trends a complex between the agricultural and consumption patterns that shape outcomes across different the MRD the in and prevalence all regions in This from improved economic as well as the production to the average per capita in the MRD from to the sharply from to while the improved from 40% to income and improved The MRD of vegetables, fruits, and aquatic including such as and This agricultural not only household income but also dietary and food through increased access to both food and The leading in aquatic into consumption the average per capita daily consumption of aquatic foods in the MRD the in the Additionally, the proportion of with access to the increased from in to in and access to health and food This from has better health and nutrition outcomes in the et al., However, the MRD is also the the rise in overweight and This a between the and households the and of foods that the agriculture most food produced is for and household consumption, to foods significant between and household dietary et al., 2017). This gap has with particularly in sugar-sweetened and consumption, contributing to a risk of overweight and obesity across all with findings by et that in Vietnam have higher of overweight due to greater consumption of processed foods and Meanwhile, face and economic constraints that their to a diet and contributing to the increasing prevalence of overweight as This pattern is with previous that urbanization, dietary shifts more foods, and et al., 2023). The with health the in the MRD to regions et al., et al., study the from the as the for food the dietary diversity, with consumption patterns and This be to agricultural that quantity over in and a food system food and poor dietary with food to agricultural the that food The increase in production, particularly between 2010 and 2014 a by to national food of Vietnam, and in and at which to consumption expanded and the from two to three per by in and production agricultural enhanced while in for and and water availability for & 2018). The in 2014 with such as and the of during that However, the period from to production as and linked with increasing to trade and trade 2019). the same the in households and the increase in households with to the a to agricultural production, particularly the in the on production has a in the of the food systems, as the MRD is one of the regions over the that food production is with household such as et al., et al., and et al., with findings from studies in In addition, studies from various countries have that a focus on production to dietary and to dietary patterns et al., et al., the consumption of foods, particularly items, with a on foods, has to the rise in food consumption in the region. the availability of from household production in the MRD consumption for all the in intake and the consumption of and that the of household diets addition, the change in consumption patterns in the MRD diversity, and foods also be linked to in et al., et al., household including income et al., et al., et al., household size et al., et al., and et al., as well as of with findings from studies MRD one of the most regions in Vietnam, by the proportion of household with or to higher of sugar-sweetened beverages and foods et al., 2018). It also limits to and that could both productivity and outcomes. This in the of and nutrition a more and food between high food production and poor outcomes is a across including the Delta of & the Mekong Delta et al., and the of and et al., 2017). these regions are as their they from a of dietary and diversity, to the on of climate change, environmental and migration (Headey & et al., et al., 2024). This the need for a approach that focus on both diets through production or and to food and nutrition of changes in the food the global from to in food and development & et al., 2019). These changes all food system consumption, and their outcomes. The with and of five in diet system and of (Reardon & The of food all five as in the study The the increase in foods, processed foods, and sugar-sweetened transformations are through agricultural land use changes, and the to three with the transformation supply and while not comprehensively in this is also through the rapid of supermarkets and as well as the of and et al., the of the MRD and their the study that the food system transformation in the MRD and faces significant in the The by similar to regions in and et al., has both and The in production and their linked and a significant to the of and practices et al., 2023). the a in food and nutrition including the supply of and produced foods 2020). The increasing of food production and the expanding of how traditional, small-scale interact with and integrate into supply et al., to influenced by economic development, urbanization, and This a a diets with higher of processed and foods, while diets rich in and aquatic products could be by health and is climate change, which influence food system in the Mekong and rising are to agricultural productivity as well as and migration that influence food consumption et al., et al., and of food system changes in MRD the of integrated that the complex system et al., et al., The both the of agricultural on production and the need for a from to and and by food and food system integrated of production production integrated systems, and to development, nutrition and These be within National for and the National on and while through dietary sugar-sweetened and comprehensive nutrition These not food system in but be integrated across all food system components to the between production and study has on data and not food production and consumption particularly food and The of data across limits analysis for all food system while the study the research not between and outcomes. The of food system transformations it to the of the regional-level focus significant and within the diverse income and provincial more detailed within this we have not analysis of supply transformations due to data and the of the findings research gaps in that production over comprehensive system outcomes. research examine how sustainable production practices health outcomes and dietary productivity to on integrated agricultural the of nutrition in addressing of while how production household consumption To scale these impacts, research focus on examining that and in sustainable food production, and sustainable food production and consumption to regions across Vietnam and Southeast communes are as within or regions, from economic and These often face such as and This is based on of the from 2010 and for each nutrient is as follows: et al., 2023).