Consumer Nutrition - an overview (2023)

5.2 Consumer nutrition environment

There have been fewer studies of the consumer nutrition environment than the community nutrition environment and the findings of these studies are mixed. There is some evidence in the US for poorer availability of healthy products in low income and ethnic neighbourhoods than more affluent neighbourhoods. The evidence from other developed countries is weaker and less consistent. The evidence for neighbourhood disparities in price is varied across all countries with findings showing both cheaper and dearer prices in disadvantaged neighbourhoods. While few studies have assessed neighbourhood differences in quality of food, the evidence consistently showed poorer quality produce in more deprived than more affluent neighbourhoods. Moving to consider how consumer nutrition environment factors relate to diet, the literature shows some evidence for associations between price and availability and dietary intake in the US. The literature and evidence from other developed countries is limited and weaker. Changes in the price of less healthy foods, particularly sweet carbonated drinks, have shown that as price increased intake of these foods decreased. This relationship was differentiated by ethnicity and income with the strongest associations observed in blacks and low-income earners. Local availability of healthy foods was also shown to relate to poorer dietary patterns but not better dietary patterns. This finding suggests that individuals with poorer diets may be disproportionately affected by their neighbourhood food environment than individuals with better quality diets. This disparity may result from being less likely to have access to a private car to travel to other shopping opportunities or more likely to keep daily activities to a more localised space (Coveney and O’Dwyer, 2009).

5.9 Policy measures to directly influence consumer malnutrition

What about food policy changes to boost consumers’ nutrition and health? The dietary transitions that occur with income growth and urbanization are substantial (Masters, Finaret and Block 2022). They are influenced also from the supply side by transport and infrastructure improvements that lower costs of trading fresh food, and by productivity growth along the food value chain right through to retail supermarkets. Yet despite the rapid penetration of fresh fruit and vegetable consumption in both rural and urban areas of low- and middle-income countries, undernutrition and obesity continue to coexist as major health concerns in those countries (Reardon et al. 2021).

Both types of malnutrition, as well as micronutrient deficiencies, can be influenced by government policies. Specific taxes on ingredients such as sugar and low-quality oils are being introduced in numerous countries (Calcott 2021). Other interventions include labelling regulations, restrictions on advertising and marketing of unhealthy products such as sugar-sweetened beverages, and nutritional education programs, e.g. for mothers and school children. Interventions that simultaneously target both undernutrition and obesity could be expanded, because of the common drivers of these often-coexistent forms of malnutrition (Hawkes et al. 2020).

4 Discussion

Though structural economic models have been extensively used for predicting demand and supply, their role in fulfilling nutrient requirements have not been discussed sufficiently in literature. Most of the policy discussions in India address nutrition, consumer demand and production separately, without exploring the complex interlinkages amongst them. The study aimed to bridge this gap, although focusing on the interlinkages at a national level. In India, the diets vary by region and socio-economic status. For example, wheat-based diets are primarily consumed in northern and western states of India whereas rice dominate the diets consumed in southern and eastern states of India (Rao et al., 2018). The regional variation in dietary pattern is also reflected in the risk of quality protein deficiency which varied from 19% to 49% amongst different Indian states (Minocha et al., 2017). Hence, not only a national level increase in availability of quality foods is important but it's regional distribution commensurate to the risk of quality protein deficiency in different states is equally important. Accessibility plays a significant role in this context. Similarly, there is also a variation in dietary patterns by rural and urban sector and by income levels. Another critical factor is the access to public distribution system (PDS) which plays a major role in driving the demand for cereals. We do not specifically segregate the demand for different groups and leave this investigation for future work. However, in general, the Indian diets as well as public procurement and distribution programmes remain heavily focused on cereals (Swaminathan et al., 2012). These offer relatively low quality protein, limiting for lysine, and the digestion and absorption of cereal protein may not be complete. Hence, it is very important to move the discussion of protein beyond its overall density in food and its amino acid score, and consider the issue of bioavailability.

There is sufficient evidence on the effect of quality protein, particularly animal protein, on linear growth of pre-school and older children (Neumann et al., 2007). Equally important is the role of other foods, such as fruits and vegetables as well as a balanced diet in providing various micro and macro nutrients (Minocha et al., 2018). Promising interventions aimed at improving nutrition through a balanced supplementation with appropriate energy-protein ratio and micronutrients have proved to be effective in prevention of adverse perinatal outcomes (Bhutta et al., 2013). More evidence from large-scale programs is needed to better understand the effect of poor quality of diet and the role of proteins, especially as India is on a health transition combating emerging public health problem of non-communicable diseases such as diabetes, hypertension and cardio-vascular diseases.

Yet more precise data and methods are required to evaluate the role of processed food, especially the ones which are packaged as having ‘high protein’, the demand for which is expected to increase amongst urban consumers. The available data does not segregate between good quality processed food and junk food. Data challenges also exist in predicting availability from production. There are few studies that explain the gap between production and availability, especially in the context of animal protein.

Moreover, the production predictions generated in the present study do not consider the role of climate change, which is expected to adversely affect both the yield and nutrient availability of crops (Myers et al., 2017). In such a scenario, the production predictions are underestimated, and greater efforts to adapt to climate change and increase in production may be required. This is particularly relevant for a developing country like India where the vulnerability to climate change may be greater because of both widespread malnutrition and a greater economic dependency on agriculture. Elevated atmospheric CO₂ concentrations may cause up to 5.3% of losses in dietary protein in India and push an additional 53 million people at risk of protein deficiency by 2050 because of decreased protein content of crops grown under such conditions (Medek et al., 2017). Further, increasing temperature and variability in rainfall is expected to alter the yields for rice and wheat, which are the major sources of protein in Indian diets (Gupta et al., 2012; Ninan and Bedamatta, 2012; Pattanayak and Kumar, 2014). In this context, a push in the production of pulses not only makes sense from a nutrient perspective but also because it is more resilient to climate change (Rao et al., 2018). For animal source foods, much of the discussion has been on it's role as a source of methane emissions, however, there is evidence to suggest that heat stress negatively effects the productivity as well as the reproduction of livestock (Sirohi and Michaelowa, 2007).

The results of the economic demand and supply models presented in the study were externally validated. The production predictions for food grains of the present study are comparable with the advanced estimates (different from final estimates) released by the Government of India for 2016–17 (Government of India, 2017). With the exception of wheat, the demand predictions of the present study are very close to the predictions provided by Planning Commission (Government of India) (Indian Agriculture Research Institute, n.d.). In addition, the predictions for both food grains as well as milk and egg, fish & meat were comparable with other studies (Kumar et al., 2016). The results of the nutrient demand model were compared with the ICMR recommended balanced diet (National Institute of Nutrition, 2011) to validate the findings. The total food required per the recommendation, converged with the sum of actual intake of protein foods and the calculated additional required intake to reduce the population at risk.

In conclusion, the predicted availability of quality protein foods, is not expected to be enough to reduce the population at risk of deficiency to less than a notional target of 5% by 2026. In other words, deficiencies will continue to exist in India in coming years. This assumes that per capita consumption of cereals stays constant. If cereal consumption increases, protein requirements might be met, but this is at the cost of a positive energy balance and the risk of obesity in the population. Moreover, the results from economic demand model suggests that if these high-quality protein foods were indeed available, the gap between their nutrient demand and economic demand will shrink over the years simply due to economic growth. EFM is an exception; although there are cultural barriers to the kind of meat consumed, efforts are required to increase its demand other than its availability. Per NSSO 2011–12, almost 80% of meat is consumed as fish and chicken whereas smaller amounts of goat meat, buffalo meat and pork are consumed.

Overall, it is the supply side of high-quality protein foods rather than demand that needs attention of the policy makers for fulfilling the protein requirements of the population. As a start, this requires the focus of agricultural policies to shift some grain production from cereals to pulses. In parallel, government food programs should also aim at shifting subsidy-based supply from cereals to quality protein foods.

Although there are strong cultural preferences for the type of cereal consumed, a shift within the type of cereals, from rice to wheat and millets, could also be useful. Wheat and millets have a comparatively better nutrient content that rice, although the bioavailability of nutrients in millets has not been much explored. However, the economic demand model presented in the study estimates a negative income elasticity of other cereals which primarily comprise of millets. This is not surprising as millets are culturally considered as a poor man's crop and its consumption has seen a decline over the years. This trend could potentially change with the recent policy push for millets in the PDS. Additionally, it also has a potential role amongst urban consumers. The availability of new products in the market such as “Multi-grain Atta”, “Ragi based dosa batter”, etc indicate that millets in a mixed form with rice and wheat have been embraced by health-conscious consumers residing in the cities.

Even though long-term policy measures should be aimed at fulfilling nutrient requirements through whole foods, other short-term options are available. The limiting indispensable amino acids (IAA in these diets is lysine), and the fortification of cereals with lysine can improve the amino acid score of the cereal protein, at least in the short run. This might also be feasible from a cost angle. The improvement of protein quality in food supplies through fortification with specific amino acids has already been tested; for example, the feeding of children with severe acute malnutrition with ready to use therapeutic food where a soya, maize, and sorghum mix was enriched with amino acids and was found to be as efficacious as a milk-peanut based food (Bahwere et al., 2017).

Similarly, a complementary cereal-legume food fortified with lysine (0.75%) is being tested for its efficacy on promoting growth in under 5 children in Ghana (Ghosh et al., 2014). The important consideration in these amino acid fortification strategies is not to cause amino acid imbalances (Kurpad, 2018). Further, there is a potential role of biofortification to improve the protein quality of cereals through conventional plant breeding, and progress with respect to development of quality protein maize and high protein polished rice has been made in this context (Neeraja et al., 2017). If successful, such strategies could be targeted towards the most vulnerable groups through Public Distribution System (PDS) or Mid-Day Meals (MDM). State level interventions can also be made but these need to be made strategically. Currently, amino acid fortification is a second alternative, with primacy going to the provision of diverse diets.

1 Introduction

The global prevalence of overweight and obesity is increasing, with rates approaching or exceeding 1 in 4 adults in numerous continents (International Food Policy Research Institute, 2016). The health, social and economic burden of obesity is well recognised (Dobbs et al., 2014; Kopelman, 2007; Reilly et al., 2003). In the UK, average intakes of sugar, saturated fat, and salt also exceed UK recommendations, and it is estimated that only 9% of 11–15 year olds and 29% of adults meet ‘five a day’ recommendations for fruits and vegetables (Public Health England, 2014). Strong calls have been made by numerous national and international organisations for policymakers to take robust action against obesity, and to help improve nutritional behaviours more generally (Government Office for Science, 2007; Institute of Medicine, 2012; World Health Organisation, 2016).

One area that has recently received attention from researchers and policymakers alike is the ‘retail food environment’ (RFE), and the link this may have with health and obesity-related behaviours. The RFE is characterised both by the ‘community nutrition environment’ (the local opportunities to acquire food) and the ‘consumer nutrition environment’ (the environment within and around food outlets (FO), comprising characteristics such as the price, acceptability and variety of food) (Swinburn et al., 2013). The concept that the RFE might be a driver for obesity is enticing, particularly to policymakers, because it suggests that it may be possible to transform environments from ‘obesogenic’ (i.e. promoting excessive energy intake, making obesity more likely) towards ‘leptogenic’ (i.e. deterring excessive energy intake through better access to healthful foods and/or fewer opportunities to obtain unhealthy foods).

Measures of the RFE are central to understanding its links with health and obesity. RFE measures broadly fall under three categories: (i) perception measures, which assess concepts like residents’ perceptions of the quality and availability of food provision; (ii) audit measures, which generally assess characteristics of the ‘consumer nutrition environment’, such as the variety and price of foods within an outlet; and (iii) Geographic Information Systems (GIS) measures, which measure the spatial accessibility of FO (see e.g. Kelly et al. (2011), Ohri-Vachaspati and Leviton (2010) and Caspi et al. (2012) for a review). These measures are typically used in isolation, although some studies incorporate multiple measures (Rose et al., 2010).

To date, GIS measures have been by far the most commonly employed. For example, a review by Caspi et al. (2012) examining the associations between the RFE and diet reported 68% of studies used GIS techniques. GIS techniques are also widely used by Town Planners and Local Authorities/Government Agencies in developing policy and making planning decisions (Glanz et al., 2016), making GIS-based research particularly relevant to policy development.

Several reviews have highlighted considerable heterogeneity in the methods used in GIS-RFE research (Charreire et al., 2010; Cobb et al., 2015; Forsyth et al., 2010; Ni Mhurchu et al., 2013). This heterogeneity makes the collation and interpretation of research findings difficult, and hinders effective translation of research into practice. While most authors acknowledge these limitations, an absence of best practices means the problems look set to persist. With such diversity in methods, accurate and transparent reporting is essential. However, various important methodological decisions are often insufficiently reported or omitted (including work by current authors).

Very little guidance exists to support authors in reporting GIS-RFE methods. Forsyth et al. (2006) propose a general framework for reporting GIS-based measures of the built environment, which calls for detailed reporting of the constructs measured, the GIS methods used, and any questions that arose during the measurement process. While this framework is useful, it is relatively general and does not consider issues specific to RFE measurement.

In view of the above, this paper seeks to (i) identify common dimensions of methodological diversity across GIS-RFE research, (ii) review the impact of different methodological choices, and (iii) highlight areas where reporting is often insufficient. On the basis of this discussion, the Geo-FERN (Geographic Information Systems Food Environment ReportiNg) checklist is proposed. Adoption of the reporting checklist will facilitate better reporting and critical evaluation of methods leading to a greater understanding of the links between the RFE, health and obesity, and improved translation of research evidence into practice. It should be noted that it is not the aim of this article to suggest ‘best practices’ for GIS-RFE methods; there is insufficient evidence on which to make such recommendations, and best practices are likely to vary depending on the specific study design and research question. However, it is hoped adoption of the Geo-FERN checklist will, enable appraisal of methods on a case-by-case basis.

4 Discussion

Measuring the nature of local food environments is important, not only to determine the types of foods that are available for purchase and preparation within the home but to assess the types of products available when consumers are making food purchasing decisions for consumption outside of the home. This literature review has identified that there are a range of available data collection methods and valid and reliable measurement instruments and indices that can be used to measure different components of local food environments.

This review has identified a number of matters that require further consideration and attention in future research that assesses the nature and extent of local food environments. Firstly, there is a need to integrate information on the types of food outlets that are available, as well as the food products sold within these outlets. Such a multi-dimensional approach to measuring food environments has demonstrated potential for a more nuanced and comprehensive assessment of these settings (Rose et al., 2010). Secondly, for measures of the community nutrition environment, the direct observation technique has been found to be a more robust method of identifying food outlets than commercial business listings (Bader et al., 2010). Thirdly, in assessing the consumer nutrition environment, measurement tools can capture a range of environmental stimuli faced by consumers that may affect their food choices, which have been grouped here according to product, price, placement and promotion. Saelens and Glanz (2009) noted that such measures should include: the relative healthiness of foods and beverages available; facilitators and supports for healthy eating; barriers to healthy eating; pricing; placement and prominence of products, and signage and promotion (Saelens and Glanz, 2009). These constructs were included in varying degrees in the survey tools identified in this literature review and can be used as a basis for future research. These measures of the consumer nutrition environment can be applied to both food outlets in communities or neighbourhoods as well as food outlets within organisational settings.

Further, to allow for the collection of such detailed information on food environments, research has shown that a selection of indicator food and beverage products can be measured rather than all items sold within outlets. Such a selective approach to data collection would add specificity and time savings (Glanz et al., 2007). Logically such measurements should focus on dietary factors related to major nutrition-related chronic diseases, including overweight and obesity, diabetes, cancers and cardiovascular disease (Glanz et al., 2008). Typical food items that have been used in surveys have included fruit, vegetables, milk, meat and alternatives, as well as a range of less healthy items, such as sugary beverages, salty snacks and confectionery. Importantly, it has been suggested that overly complex measurement tools, which capture perhaps unnecessarily detailed information on food environments, may not facilitate an understanding of these settings nor clearly identify potential intervention points (Rose et al., 2010).

It is thus important to present findings in a way that is readily comprehensible. For example, the development of composite or integrated indices, which take into account a range of environmental factors, may assist in data reduction and in making simple comparisons between different outlets and locations. Using such an approach may also assist in balancing the complexity of food environments and ensuring that the more salient features are captured and transparently reported (Rose et al., 2010).

As well, in some instances, the use of simple short questions relating to consumers' purchasing behaviours can be useful to assess the products commonly purchased from a setting or an outlet, or to assess the frequency of purchases. These questions may provide a useful adjunct to objective measures of food availability to determine any association between availability and consumer behaviours (Nickelson et al., 2010).

For practitioners working in this field, a first step is to consider what data are already available that can be summarised to present an overview of the local nutrition environment. If resources are limited, the use of a checklist approach, which captures a range of indicator foods and conditions may be most appropriate, while a number of more detailed tools are available to provide descriptions of particular nutrition environment constructs with additional resource inputs. However, it is always important to adapt and test any instrument to ensure its relevance and reliability when used in a new environment.

Finally, in devising new instruments to measure local food environments based on the research to date, rigorous pretesting in the formative stage of instrument development will ensure that the tool is feasible, highly reliable and repeatable, with good face validity. Reliability will need to be assessed and can be undertaken by repeated measurements of the same location on the same day by different researchers (inter-rater reliability). Repeatability also will need to be determined and can be assessed by revisiting and assessing locations over time. These measures will assist in confirming that a new tool can be used consistently by multiple observers, and if cross-sectional information is adequate or if multiple measures are needed to capture changes in stocking or available promotions. Face validity can be assured by basing the measurement tool on previously tested instruments and verifying that all important constructs are being measured. While inter-rater reliability was reported for some of the instruments and techniques identified in this review, more transparent reporting is required for other measurements. Regardless, seasonality is likely to affect the availability, price and quality of some foods (Glanz et al., 2007).

This literature review has some limitations. As this review sought to identify a range of measurement tools and methods to describe different food environment constructs, we did not provide a critique of all possible measurement instruments and indices available. As well, this review did not describe the use of geocoding, which may also be used to analyse spatial access to food outlets in relation to other geographic coordinates, such as neighbourhoods. Rather, other measures of food outlet availability and density have been presented.

Many of the studies measuring food environments to date originate from the USA and relate to food environments in or around residential settings. Research from a wider range of settings, including the food environment that people are exposed to in their workplaces, recreation areas, and during their commute to work and recreation is required to determine the nature of food environments more broadly. Accurate measures of food environments are necessary to inform policy making, particularly in relation to the placement or zoning of food outlets and food labelling or nutrition information requirements. Importantly, descriptive information relating to the nature of food environments also assists in identifying intervention points to improve communities' access to, and the availability of, healthy foods.

3.1 Pandemic-driven consumer behaviour (RQ1)

Health-related behaviour. Not surprisingly, the pandemic causes consumers' fundamental concerns on personal health, which are integrated into consumption activities, leading to health-related behaviours [20]. Among these behaviours, consumers' adoption of preventive behaviour (e.g. self-isolation, social distancing and mask usage) is most frequently examined (n=6). Besides, consumers are more conscious about consumption of (un)healthy food (n=6) and subjective wellbeing and healthy lifestyle (n=4). For example, Aksoy, Kabadayi and Alan [21] examined consumers' nutrition intake during the pandemic focusing on factors such as health consciousness, social influence and COVID-related fear emotions. Meanwhile, Liu, Lv and Tang [22] noted an interesting behaviour of quantified-self by consumers (e.g. record calorie intake and running mileage) in response to the pandemic. In addition, the pandemic also drives a massive trend of service digitalisation, and our review identified consumers’ usage of digital healthcare tools (n=2) as one exemplified behaviour.

Abnormal buying behaviour. Perhaps, the most sensational impact of the pandemic on consumers is the resultant abnormal buying behaviour. Under different terms, researchers have examined consumer behaviours of panic buying (n=15), impulsive buying (n=6) and compulsive buying (n=1). The panic buying behaviour is largely driven by consumers' basic needs of security and safety (e.g. purchase food and medical supplies in a large quantity), whereas impulsive/compulsive buying is more subtly related to consumers' needs for psychological assurance [2,8]. Nonetheless, all the exemplified abnormal buying behaviours are associated with consumers’ emotional state of fear and anxiety due to the uncertainties during the early pandemic stage, building up an unusual mentality of hoarding and stock-up (n=3).

Technology-related behaviour. The pandemic also leads to some unexpected positive changes in consumer behaviour, such as the expedited technology acceptance by consumers. Within the category of technology-related behaviour, consumers' usage of multi-channel shopping technologies (n=10) and contactless delivery technologies (n=6) has frequently been examined. Although these technologies are not new to modern shoppers, the pandemic has brought new considerations, such as fear appeal, risk perception and social norm, into consumers' response to technologies [23–25]. In addition, other technologies that facilitate consumers’ social-distanced lifestyle, such as digital financial tools (n=5), online connection tools (n=2), digital healthcare tools (n=2) and virtual tourism tools (n=1), have also been investigated.

Information-related behaviour. Related to the trend of digitalisation is consumers' behavioural responses to massive information during the pandemic. These behaviours are categorised as information-related behaviour. To illustrate, our review identifies three studies that examine consumers’ search and processing of health-related information. Consumers experience emotional stress and cognitive overload when facing an explosive amount of information mixed with fake news and exaggerated tones [74,76]. Consequently, consumers also demonstrate diverse information-related behaviours, such as sharing fake news (n=1), relying on social media for information (n=1) and informational conformity (n=1).

Leisure-related behaviour. The tourism and hospitality literature also contribute considerable research on pandemic-driven consumer behaviours. These behaviours are referred to as leisure-related behaviours. Research attention has been paid to consumer behaviours of visiting restaurant (n=4), hotel (n=2) and retail store (n=2), travelling by flight (n=2), cruise (n=2) or travelling in general (n=2). In addition, the pandemic has given rise to new forms of leisure activities, such as virtual reality (VR) tours (n=2) and short-distance one-day tours (n=1). Given the dramatic impact of COVID-19 on the tourism and hospitality industry, consumers are more concerned about safety issues, precautionary measures and informational uncertainties [79,84,85], which were unconventional considerations in consumers’ leisure activities before the pandemic.

Prosocial behaviour. The last category concerns another positive change in consumer behaviours, that is, prosocial behaviour. Our review suggests that, despite the disruptive nature, the health crisis promotes some consumer behaviours that are socially or environmentally desirable. This is probably due to the economic, environmental and social disruptions commonly felt by the entire population that ignite the innate goodness of human being. In this regard, Zwanka and Buff [35] suggested that the pandemic leads to a shift in consumer behaviour from care for oneself to care for the greater good. For example, consumers demonstrate support for local/national brands (n=4) and brands that ethically respond to the pandemic (n=1). Moreover, pro-environment actions (n=3), charitable actions (n=2) and voluntary advocating behaviours of preventive measures (n=2) have been reported.