How Livelihood Change Affects Food Choice Behaviors in Low- and Middle-Income Countries: A Scoping Review

Livelihoods have changed dramatically over the past decade in low- and middle-income countries (LMIC). These shifts are happening in tandem with shifts in individual and household food choice behaviors. This scoping review aimed to identify and characterize mechanisms through which livelihood changes could affect food choice behaviors in LMIC, including behaviors relating to food production, acquisition, preparation, distribution, and consumption. A literature search was conducted using 4 databases: PubMed, PsycInfo, AGRICOLA, and Embase. The search was further enhanced by expert solicitations. Studies were included if they measured or focused on a livelihood change, described or assessed a change in ≥1 food choice behavior, and focused on LMIC. Studies were excluded if they focused on migration from LMIC to a high-income country. Of the 433 articles that were identified, 53 met the inclusion criteria. Five mechanisms of how livelihood change can affect food choice were identified: occupation, locality, time, income, and social relations. Changes in occupation altered the balance of the availability and affordability of foods in local food environments compared with individual food production. Changes in location, time use, and income influenced where food was purchased, what types of foods were acquired, and how or where foods were prepared. Additionally, changes in social relationships and norms led to expanded food preferences, particularly among urban populations. Time limitations and higher discretionary income were associated with consumption of ultraprocessed foods. Understanding the relationships between the changes in livelihood occuring in LMIC and food choices of households in these countries can inform the development of policies, programs, and other actions to promote sustainable healthy diets and planetary health.


Introduction
What and how people eat is central to all cultures and intertwined with the livelihoods of the world's populations.Over the past century, livelihoods globally have undergone transformations driven by technologic changes, social and economic factors, environmental influences, and individual choices, with rapid shifts over the last decade in Africa, Latin America, and Asia [1,2].Although livelihoods can be defined broadly as a means of support or subsistence [3], for the scope of this article, livelihoods are defined as "activities that individuals participate in to acquire resources for securing necessities of life and to meet material and nonmaterial goals, including the maintenance or attainment of social status or membership in a particular group" [4].
People maintain their livelihoods through participation in formal and informal labor markets, as well as forms of passive resource-generating activities [5].Labor markets consist of diverse employment arrangements characterized by particular schedules (e.g., shift work, full time, part time, and seasonal work), duration of work (e.g., fixed-term project or task-based occupations, casual or seasonal labor), geographic location (e.g., rural compared with urban, relocation patterns), compensation structure (e.g., hourly wages, daily wages, annual salaries, and piece work), employer-employee relations (e.g., supervisor-subordinate, self-employment, subcontracting, and temporary agency), and sector of the economy (formal or informal) [6].In low-and middle-income countries (LMIC), over 75% of eligible adult workers are employed in the informal sector (e.g., occupations not subject to labor regulation, taxation, social protection, or benefits) [6].Some livelihoods involve precarious conditions, including, intermittent work, more hours per week, lack of occupational safety and social protection, and limited free time [7].Furthermore, work and income instability, particularly in the informal sector, disproportionally involve communities in poverty and add to the challenges faced in daily living.
Livelihoods change when individuals shift or adapt their means of supporting themselves and their households.This includes changes for people historically involved in food production.Globally, ~1.23 billion people are employed in agrifood systems and 3.83 billion live in households linked to agrifood system-based livelihoods [8].Employment in agrifood systems has declined in Africa from 60% in the early 2000s to 53% in 2019 and from 55% to 35% in Asia over the same period [8].In many LMIC, recent large-scale livelihood changes have involved movement from on-farm to off-farm activities within food value chains, such as food processing, packaging, and distribution, or shifts out of food-related livelihoods entirely, or nonfarm activities such as mining or construction [9].Livelihoods have also shifted from rural to urban areas, including the growth of secondary towns and cities where more employment opportunities are available [10].Additionally, more females are seeking work opportunities outside the home, often alongside continued responsibility in domestic care and social reproduction [11].
Livelihood changes can dramatically influence the daily lives of individuals and households, impacting diets and human and planetary health through major alterations in individual and family food choice behaviors to changing food environment [12][13][14][15][16][17].Food choice is the decision-making process, both conscious and unconscious, through which individuals produce, acquire, prepare, distribute, and consume foods [18,19].What, how, and why people eat the way they do must be understood within the biologic, economic, cultural, social, and environmental context in which the decision(s) are made [18][19][20].Prior studies have demonstrated relationships between livelihood change and food choice behavior, but there is a lack of evidence on the mechanisms through which livelihood changes affect individual food choice behavior [21].Livelihood changes may be forced or voluntary and are related to the power or agency of the people involved, resulting in both positive and negative outcomes.Forced changes are often to minimize risk from external factors (e.g., climate change, war, and disease), whereas voluntary changes are usually taken to invest in future opportunities [22,23].These 2 reasons can have different implications for how households make food choice decisions.
This scoping review aimed to identify and characterize the mechanisms through which livelihood changes could affect food choice behaviors, including behaviors relating to food production, acquisition, preparation, distribution, and consumption.The review focuses on individual-and household-level changes rather than changes in food systems.Understanding the relationships between livelihood change and food choice can inform the development of policies, programs, and other actions to promote sustainable healthy diets and planetary health.

Methods
The Drivers of Food Choice (DFC) Program was established to facilitate, synthesize, and disseminate research to provide a deeper understanding of what, how, and why people eat the way they do in Africa and Asia [24].Over a 5-y period, the DFC program funded 15 projects in 10 countries, using a wide range of methods and led by diverse investigators from 38 collaborating institutions.Affiliates of the DFC program with expertise in food environments, food systems, and food choice research convened to identify priorities for evidence synthesis during the Agriculture, Nutrition, and Health Academy Week in Hyderabad, India, in June 2019.Livelihood characteristics were consistently identified as central to food choice.Findings from the DFC program portfolio guided initial conceptualization of the relationships between livelihoods and food consumption, and also informed the scope of this literature review [25][26][27][28][29][30].

Search strategy for the systematic scoping review
A systematic scoping review was performed using the PRISMA-ScR [31].To identify potentially relevant articles, the following bibliographic databases were searched from November 2021 to July 2022: PubMed, PsycInfo, AGRICOLA, and Embase.Search strategies were drafted by coauthors (EK, KR, SS, CB) and refined through team discussion (Supplemental Table 1).The electronic database search was supplemented by expert solicitation (i.e., DFC technical advisory group, external DFC grantees, and well-known experts in the field) of articles from the DFC program and Google Scholar to expand the range of research articles and gather diverse perspectives.

Inclusion and exclusion criteria
Articles were included if they met the following criteria: 1) included 1 LMIC defined by the World Bank [32], 2) measured or focused on a livelihood change, 3) described or assessed a change in 1 food choice behavior (production, acquisition, preparation, distribution, and consumption), 4) peer-reviewed, and 5) published between 2015 and 2022.Quantitative, qualitative, and mixed-method studies using primary and secondary data were included.The year 2015 was used as a starting point because this is the year when the research community recognized the need to better understand food choice in LMIC, including through initiation of the DFC program [18,33,34].Although the DFC program focused on Africa and Asia, articles from Latin-America were also included if they were set in countries defined as LMIC by the World Bank [32] between 2015 and 2022: Belize, Bolivia, El Salvador, Guatemala, Haiti, Honduras, and Nicaragua.Articles were excluded if they did not meet the criteria above or focused on migration from an LMIC to a high-income country.

Data screening
Articles were screened independently by 3 coauthors (EK, KR, and SS) for eligibility.Title and abstract screening were followed by full-text article screening.Duplicates were removed.After screening, the working group discussed the final articles for analysis (Figure 1).

Data charting
A data-charting form was developed for data extraction.The following contents were extracted: title, first author, population, description of the livelihood change, mechanisms for how livelihood change affects food choice behaviors, description and dimension of the food choice behavior change (production, acquisition, preparation, distribution, and consumption).In this article, we define production as decisions about what and how food is grown or bred (for livestock, dairy, poultry, and fish) through agriculture, pastoralism, and/or aquaculture [35].
Acquisition is defined as food procurement through wild food gathering, production, market purchase, or through the social economy (e.g., bartering, gifting, direct monetary transactions) [36].Preparation of food is the application of techniques (e.g., cleaning, chopping, and heating) to a combination of several foods or raw food materials to create a meal [37].Distribution of food is how food is shared within a household among members [38].Consumption is the process of ingesting food and involves eating episodes, including characteristics such as location and with whom [19].
The mechanisms of how livelihood changes could affect food choice behaviors were identified through qualitative thematic analysis of included studies [39].Included articles were organized and independently charted by food choice behavior and reviewed to identify the mechanisms of livelihood change impacting each behavior.Coauthors (EK, KR, and SS) discussed results and iteratively updated the data-charting form.Additional coauthors were consulted when further clarification was needed.To ensure scientific rigor and reduce subjective bias, coauthors overlapped in the sections reviewed and reconciled interpretations [40].When articles included >1 food choice behavior, each behavior was charted separately before combining all articles into 1 matrix.The final tabulation included each article once (n ¼ 53), indicating which food choice behaviors were represented.

Occupation
Occupations shift when individuals transition job or profession, or engage in multiple jobs (diversification), affecting production, acquisition, preparation, distribution, and consumption of foods.Of the 53 articles that were reviewed, 37 reported a change in occupation, resulting in a change in production (n ¼ 17), acquisition (n ¼ 21), preparation (n ¼ 12), distribution (n ¼ 9), and consumption (n ¼ 32).Findings for production included shifts to nonfarm livelihoods being closely associated with reduced food production [27,42,50,57,88].Increased market engagement and shift to market-oriented crops changed diets.Changes in commercial agricultural engagement-such as increased livestock production in Cambodia and Zambia [48,54], aquaculture farming in Bangladesh [56], or cash crops in Vietnam, Cambodia, Tanzania, and Indonesia [28,55,61,63]led to shifts in the quantity and types of foods produced for commercial compared with household use.Without own food production, people depended on traditional and modern retailers to acquire food [27,53].With increased dependence on food markets, households changed their preservation or storage techniques, allocation practices, and dietary patterns [28,29,53,58].Overall consumption of kilocalories and diversity of foods increased when individuals shifted from traditional agricultural livelihoods to nonfarm livelihoods or self-employment [59,62,67].Shifting from subsistence agriculture to aquaculture or small livestock production increased the consumption of animalsource foods [48,49].In contrast, shifting from farming to nonfarming occupations decreased the consumption of traditional staples and animal-source foods and increased the purchase and consumption of ultraprocessed, convenience foods [28,46].
For example, in rural Indonesia, shifting from traditional agriculture or hunter-gatherer livelihoods to oil palm production shifted food choice behaviors [29].Both males and females reported allocating more time to oil palm production and less time to agricultural production compared to recent years, resulting in purchasing and consuming preprepared foods from mobile vendors.As females spent more time working outside the home, their evenings often involved performing domestic activities, such as preparing the next day's breakfast.With their new occupation, females also reported leaving their children at daycares, where workers fed their children, rather than being solely responsible for the child's food allocation.Overall, shifting to oil palm production negatively impacted food consumption patterns with an increase in unhealthy and economically unsustainable foods [29].

Locality
Locality is determined by the physical location of the workplace, which sometimes changes, such as through migration from a rural work environment (e.g., mines and farms) to an urban work environment (e.g., factories and street food stalls).Characteristics of localities have also changed in recent years with urbanization converting rural agricultural spaces into urban built environments [10].Of the 53 articles that were reviewed, 26 reported a change in locality, resulting in a change in production (n ¼ 9), acquisition (n ¼ 21), preparation (n ¼ 7), distribution (n ¼ 9), and consumption (n ¼ 23).Migration can be voluntary, in which individuals seek improved economic opportunities, or forced, due to danger, legal injunction, or external circumstances.Voluntary and forced migration typically reduced individual and household agricultural production [27,55,57,68].Voluntary migration led to increased food availability and accessibility in a new food environment, affecting food acquisition and preparation behaviors and, subsequently, diets [46].Lengthy work commutes and physically demanding occupations increased demand for mobile street food vendors near workplaces [30,44,51,60,64,66].
For example, young males in Belize who migrated to cities for informal wage work developed aspirations for modernity, having experienced different urban lifestyles and cultures; these new aspirations sometimes shifted diets away from traditional foods and toward Western or convenience foods prepared outside the home, such as fried chicken [64].Once these workers returned to their villages, they introduced urban tastes into their rural families' diets.
Migration because of urbanization impacting land use was also associated with negative changes in food choice behaviors.For example, in Ghana, urban sprawl limited arable land available for agricultural activities, resulting in lower farm productivity.This forced households to reduce food consumption and dietary diversity, ultimately increasing household food insecurity [45].

Time
Time available for food-related behaviors shifts with livelihood changes.For example, some livelihood changes require longer and less flexible working hours, or longer commutes.A change in time use was reported in 23 of the articles reviewed, resulting in changes in production (n ¼ 7), acquisition (n ¼ 16), preparation (n ¼ 10), distribution (n ¼ 11), and consumption (n ¼ 20).As individuals shifted to nonfarm and off-farm activities, less time and effort was spent on household agricultural production, decreasing individual production or increasing hired wage labor to conduct on-farm activities [27,29,30,42,55,56].With reduced time available for production, households acquired more food outside the home [27,44,46,60,64,66].Despite the perception that foods prepared from scratch were healthier, the real or perceived time required to prepare these foods resulted in the procurement of ready-to-eat foods (e.g., street foods, fast foods, and snacks) or those that were quicker and easier to prepare (e.g., instant noodles and processed meats) [46,64,66].More time spent commuting to and from workplaces and longer work hours disrupted eating and food distribution schedules and was associated with skipped meals or negative    changes in the composition and quality of meals (e.g., consuming preprepared foods) [29,44,55,65].For example, in Mozambique, females' participation in wage work in addition to farming created time constraints for food preparation and consumption, which reduced the number of meals consumed or led to repetition of the same meals at different mealtimes within a day [52].The study reported that in the past month, 43% of respondents skipped a meal because of the lack of time to prepare it and 65% reported that someone in their household had rejected a meal to avoid eating the same food for both lunch and dinner [52].

Income
Income changes refer to a shift in financial compensation when individuals or households adopt a new livelihood (e.g., wage work to self-employment).A change in income was reported in 39 articles, resulting in alterations to production (n ¼ 19), acquisition (n ¼ 24), preparation (n ¼ 11), distribution (n ¼           11), and consumption (n ¼ 34).As cash-generating activities increased, food production decreased, and people depended more on food markets [27,28,42,50,55,57,68].Affordability is a key food choice consideration, and unstable employment, particularly in the informal sector, may impact how much and what types of foods were purchased, prepared, distributed, and consumed [30].Higher incomes from livelihood change were associated with investments in not only more nutrient-dense foods, such as animal-source foods, fruits, and vegetables [12,27,47,66,68], but also energy-dense, nutrient-poor foods, such as sugar-sweetened beverages or biscuits [46,51,52,66].In West Africa's artisanal and small-scale mining sector, there was high earning potential, but the lack of guaranteed daily compensation affected food choice [30,47].In northeastern Guinea, females shifting to artisanal and small-scale mining obtained limited food from self-production and were largely dependent on markets.They often selected vendors with whom they could cultivate a relationship, enabling food purchasing on credit.These arrangements were critical for miners who faced income instability [27].Miners typically used their earnings to purchase fish, meat or poultry, vegetables, and cooking oil, but these nutritious food purchases were contingent on the productivity of the day's labor [27].Similarly, female miners routinely purchased food upon returning from the mining sites; therefore, what was prepared for the evening meal reflected the miner's daily success [30].If income was lacking, females might skip meals themselves to ensure enough food for their families.

Social relations
Social relations create dynamics around food behaviors [89].Changing gender roles and interpersonal connections, including more females working outside of the home, exposure to global media, and interactions with different cultures, may affect food choice behaviors.Changes in social relations were reported in 21 articles, resulting in a change in production (n ¼ 8), acquisition (n ¼ 15), preparation (n ¼ 11), distribution (n ¼ 12), and consumption (n ¼ 18).Females' involvement in on-farm occupations increased food production [58], whereas females' involvement in off-farm occupations decreased production [28,29,55].Females' involvement in paid labor was associated with increased decision-making power, positively affecting food acquisition and distribution, and usually improved the household's dietary diversity.However, results were mixed for food preparation practices: some studies reported no changes in food preparation practices, whereas others reported a negative shift to quicker, convenience foods [43,58].In urban areas, exposure to global eating patterns, modern mass media, and formal and informal nutrition knowledge led to changes in food preferences and habits, ultimately increasing acquisition and consumption of ultraprocessed foods and meals away from home [28,44,46,64].
For example, in rural Guatemala, mothers who shifted from homemaking to working outside the home changed their food purchases to include more meat and dairy, increasing food quantity and variety [66].Several mothers also purchased "indulgent," energy-dense foods and reported greater exposure to take-away and street foods compared with nonemployed mothers.Some mothers also enlisted their mothers or mothers-in-law for food acquisition, preparation, and distribution practices.Other females discussed challenges delegating domestic tasks to other household members, leading them to rely on convenience foods to feed their families because of time limitations and accumulation of other household tasks [66].

Discussion
This review identified and characterized the major mechanisms through which livelihood changes affect food choice behaviors: occupation, locality, time, income, and social relationships.Food production behaviors were affected by livelihood changes, particularly shifting to market-oriented occupations; this process altered the balance of the availability and affordability of foods in local food environments compared with own food production.Acquisition behaviors were heavily influenced by all mechanisms but especially by changes in income that impacted the types of foods purchased.Locality changes affected the types of foods available for purchase as well as the quantity and quality of wild and/or indigenous foods available.Changes in time use influenced when and how individuals were able to acquire foods.Preparation behaviors were sensitive to time limitations induced through livelihood change that affected the amount of time available for preparing meals, whereas changes in location altered food preparation methods.Food distribution and consumption behaviors were also heavily influenced by livelihood-induced time limitations, with the need for convenience coupled with higher discretionary incomes being key drivers of processed and packaged food consumption.Altered social relationships and social norms were a consequence of livelihood changes that led to expansion of food preferences to include nontraditional food consumption, particularly among urban dwellers.
Global changes in economic process, migration, urbanization, food environment changes, climate change, and crises such as the SARS-CoV-2 pandemic and ongoing wars have corresponded with large-scale changes in food supply chains and food choices resulting in dietary patterns moving toward ultraprocessed foods that contain high amounts of refined carbohydrates, sodium, and saturated or transfats [90][91][92].Livelihood shifts in types of occupation, work conditions, and work environments have shown some associations with increased reliance on ultraprocessed foods [93].Females working outside the home and alterations in gender roles substantially changed time use available for food choice and dietary patterns.The consumption of ultraprocessed foods contribute to poor health, including overweight, obesity and diet-related noncommunicable diseases, and environmental harms, with increased greenhouse gas emissions, loss of biodiversity, and environmental degradation from food production [94].
Time limitations appeared to contribute to reliance on unhealthy foods such as preprepared, ultraprocessed, convenience foods, leading to various negative health consequences, including the widespread adoption of unhealthy food choices [95][96][97][98][99]. Unhealthy eating patterns can result from an emphasis on convenience related to caregiver time limitations, ease in access to ultraprocessed foods, and exacerbated by further livelihood changes, with implications that are intergenerational when children's early life food exposures and diets are suboptimal.A literature review on factors contributing to childhood malnutrition found that an increase in maternal employment may increase occupational time demands, potentially limiting their ability to provide sufficient care for their children [100].This, in turn, may be associated with child malnutrition [100].
Similar to high-income countries, people in LMIC are demanding foods with shorter and simpler steps to prepare, minimal ingredients, and which can be made in advance and preserved or stored for long periods [41,46,66,101].When faced with a choice between maintaining traditional food practices or purchasing affordable foods that free up time, findings from the review suggest that many households will choose the latter because of livelihood changes.Successful campaigns to promote healthy traditional food consumption must incorporate the value of time for their target populations.For example, in India, a campaign to promote millet consumption to improve food security and health acknowledges that preparation of the grain is very time intensive and may not appeal to households with constraints on time or facilities [102].
Locality shifts related to livelihood changes are heterogeneous.For instance, 1 family member may migrate to an urban area to engage in wage work, whereas the remaining members remain in the villages and continue to maintain the family farm or other rural business [12,103].In other instances, individuals or households keep their rural residence and commute daily to an urban or periurban area for work [104].Others move their entire family unit from a rural to urban or periurban environment [23].These locality shifts may have different consequences for food choice behaviors and healthfulness of diets, particularly depending on proximity to markets and greater commercialization [105].
Out-of-home livelihood opportunities for females can increase empowerment through higher incomes, access to information and resources, and expansion of social networks.Females who shift from homemaking to employment outside the home often have more decision-making power for household food acquisition and consumption [106][107][108][109]. Less time for home food preparation, however, can decrease dietary diversity and increase reliance on ultraprocessed foods, especially if another household member does not take on some of the burden of food acquisition, preparation, or distribution.Interventions that target household, maternal, or child nutrition may be more successful if they address household food roles and shared responsibility directly.An example of an intensive females' empowerment intervention focused on improving employability of females achieved positive impacts on food security and weight status through education, self-esteem and self-efficacy promotion, and changes in gendered behaviors related to coresponsibility and self-care [101].
Recognizing the vital intersection of livelihood change and dietary health, interventions and policies aimed at effectively addressing these dynamics are of great importance.Some policies, however, can exacerbate the negative impacts of livelihood changes on healthfulness of diets.For example, policies attempting to address food safety concerns by eliminating the availability of informal food outlets (e.g., wet markets and openair markets) and building modern food outlets (e.g., supermarkets) can have unintended consequences by reducing healthy food access and compromising the livelihoods of informal sector workers those who are both vendors and consumers [63].Informal food retailers were particularly adversely affected by the COVID-19 pandemic, with closure of markets resulting in loss of income and amplifying risks of food insecurity [110].Because of the lack of proper guidance and implementation in other informal sector occupations, the issue of worker exploitation continues to present challenges, as these occupations are "under the table" and workers typically do not receive benefits, injury protection, or job security, which can hinder the ability to access nutritious foods and healthcare [7].This is the first review to the best of our knowledge to synthesize research on livelihood change affecting food choice behavior changes.The goal is for this work is to guide future research in this area, which is increasingly important with a likely increased frequency of livelihood changes in LMIC occurring with escalating climate change and other pressures.Only 3 articles in this review were from Latin America, highlighting the need for further research in this context.This could potentially be attributed to the reclassification of numerous countries in the region, which were historically categorized as lower-middle-income but have since been redefined as uppermiddle-income economies.Because of the nature of a scoping review, this review did not include an assessment of the quality of the articles reviewed.There also may be additional relevant articles that were missed because of database selection, period (e.g., restricting article publications to those published between 2015 and 2022), or exclusion of articles published in a language other than English.A scoping review was deemed more suitable than a systematic review for this study because of its emphasis on mapping the breadth of available literature, identifying knowledge gaps, and facilitating a comprehensive understanding of livelihood change and food choice behaviors.
The designs and methods of the reviewed studies provide plausible evidence that the identified mechanisms operate in a causal manner.Although many of the studies were crosssectional, many used mixed methods, important for data triangulation, and longitudinal methodologic types for which it is easier to infer causality.Most studies used a qualitative method that provided in-depth understanding of the relationship between livelihood change and food choice behaviors including the causal direction.Only 2 quantitative studies used randomized control designs, so future quantitative research that uses experimental or quasiexperimental designs would likely be fruitful, as would longitudinal research examining these mechanisms over longer time periods.
Five mechanisms-occupation, locality, time, income, and social relations-for how livelihood change affects food choice behaviors were identified (Figure 2).These livelihood changes lead to changes in individual food choice behaviors, including decreased household food production, increased acquisition of foods from markets, decreased preparation of food at home, altered household meal patterns, and increased consumption of ultraprocessed preprepared convenience foods.Understanding the changing patterns of daily life in LMIC amid livelihood changes can help ensure research recommendations for improving food choices are contextually grounded for target populations.A more detailed understanding of the mechanisms through which livelihood changes affect nutrition behaviors and outcomes could help to inform appropriate policy and programmatic levers for intervention to support vulnerable populations in achieving optimal health and well-being.

FIGURE 2 .
FIGURE 2. Indicates relationships between livelihood change mechanisms (occupation, locality, time, income, and social relations) and individual and household food choice behavior changes (production, acquisition, preparation, distribution, and consumption).

TABLE 2
Summary of relationship between livelihood change and food choice behavior changes for included studies (continued on next page)

TABLE 2
X(continued on next page)

TABLE 2
X (continued on next page)