Households with and without the presence of adolescents, probability of expenditure on food consumed away from home, according to
ENAHO 2021: a cross-sectional study

Introduction

In Peru, 42% of households consume food away from home.¹ The consumption of food away from home is an eating behaviour that is considered mostly unhealthy,² because these foods are often high in refined sugars, saturated fats and sodium, named critical nutrients.^2–4 Likewise, said foods have already shown to provide a high energy load, and can increase the weight of those who consume them.⁵ The consumption of energy-dense foods can lead to non communicable diseases (NCDs), including type 2 diabetes mellitus, hypertension, cardiovascular diseases, overweight, obesity, among many others.⁶ These diseases bring many collateral effects with them such as low productivity and incapacity in the workplace, high costs for nosocomial hospitalization, treatment and medical surveillance, and all this as a result of the social, cultural, political and economic changes that society has experienced worldwide.⁷

Currently, overweight and obesity are considered two of the most dangerous chronic noncommunicable diseases (NCDs), triggering riskier disorders.^8,9 According to WHO, In 1975, 4% of children and adolescents were overweight and obese, in 2016, this figure increased to 18%. Nearly 3.4 million people worldwide die from these diseases, and 3.8% of the global burden of disease is due to NCDs. Worldwide, obesity has increased almost threefold since 1975.¹⁰ In Europe alone, 12% and 18% of men and women, respectively, are obese.¹¹ In Latin America, the child and adolescent population suffering from obesity is estimated to be around 25%.¹² According to the Global School Health Survey, in Peru, the adolescent population who are overweight or obese is around 22.8%, and according to the National Institute of Health (INS) it is 24.2%, which means that one in four adolescents is overweight or obese.¹³

Many adolescents maintain unhealthy eating behaviours or habits, such as consuming processed food,¹⁴ and foods consumed away from home,¹⁵ with a high quantity and frequency. This acquired habit is driven by different factors ranging from biological, social, family, cultural, among others.¹⁶ These factors influence the eating behaviour and tastes of young people, children, adolescents,¹⁷ and are maintained until adulthood.¹⁸ Adolescents are an important sector of the population that is going through a period of important changes and part of these changes are dietary and therefore are in a state of great vulnerability to this situation, causing an optimal life expectancy or not.¹⁸

It should be noted that dietary behaviour has different perspectives, but food consumption patterns encompass a dominant pattern given by world globalization.^19,20 Multiple studies have shown the transition from consuming traditional foods to more and more processed foods. In the long term this has a great role in current public health problems.²¹

Nutritional and physical growth problems vary from place of origin, geographical regions of a country and between countries and some variables such as income, household wealth, unmet basic needs, economic stability, and food consumed away from home could explain these changes in adolescents.²² Populations are undergoing significant transformations in dietary and nutritional patterns as part of a process known as nutritional transition.²⁰ Peru does not escape this transition process,²³ so it is necessary and fundamental to study what is affecting the nutritional status of adolescents, since it is a very sensitive stage of life, which is immersed in biological, psychological and cognitive changes²⁴ and an accelerated rate of growth and development.²⁵

However, in recent decades the food culture has undergone certain variations,²⁶ largely as a result of globalization, changes in family patterns, lack of knowledge of parents, since they are directly responsible for the nutrition of family members, poor eating habits, fast food advertising, among others.²⁷ All these factors, alone or together, are responsible for the increase in overweight and obesity.²⁸

Due to all the above mentioned, the adolescent population has been prioritized, since they are vulnerable and receptive to their entire environment, whether it be physical appearance, beliefs and preferences, or lifestyles.²⁹ Part of these lifestyles is reflected in the consumption of food away from home.³⁰ Likewise, consumption of food away from home is one of the predisposing factors for acquiring diseases such as NCDs.³¹

In this regard, there are very few studies in our environment that investigate the food consumed away from home in households where there is an adolescent. In this study, we will compare the expenditure on food consumed away from home in households where there are adolescents and in households where there are no adolescents in the year 2021. This research has been prepared based on the National Household Survey (ENAHO) 2021. This research will only look at the expenditure factor and the demographic and socioeconomic factor, even though it is known that there are other factors such as biological, psychological, cultural, that would be influencing the nutrition of adolescents and that will not be considered in this study.³²

This research aims to compare the expenditure on food consumption away from home in households that have adolescents, as well as households that do not have adolescents in their composition. On the other hand, this research attempts to provide new knowledge on expenditure variables, as well as demographic and socioeconomic aspects as factors that could be influencing food consumption³² in households with adolescents and, above all, to improve the understanding and analysis of the problems of adolescent food and nutrition.³³

Methods

The research design is based on a quantitative approach with a cross-sectional study type because it describes a phenomenon in a population, there is no intervention and it will only be limited to measure and describe the studied population as such. This database is defined as secondary.

Results

The result of households that were included in the analysis was 4,350,171, according to the National Household Survey (ENAHO) for the year 2021, and 24245 after applying the expansion factor.⁶⁰ Households with the presence of adolescents and households without adolescents were selected and compared. Table 1 shows the average expenditure on food consumed away from home in households with and without adolescents per month. It also shows that expenditure is higher in households with the presence of adolescents, with an additional 57 Peruvian Soles than in households without adolescents; likewise, it shows the proportion of households and the standard deviation represented and the total population of households with at least one member between the ages of 10 and 19 years, which is represented by 43.92%. On the other hand, households without adolescents represented 56.08%. The results were obtained with the application of the expansion factor (Table 1).

Table 2 shows the difference in means, in households without and with adolescents; the result indicates that the averages are statistically different.

Household with adolescents

_subpop_1: tipo_hogar_adolescent = Household without adolescent

_subpop_2: tipo_hogar_adultos = Household with adolescent

Ho: [GAFH]_subpop_1 - [GAFH]_subpop_2 = 0

The highest expenditure is made by households with adolescents, the percentage of the population in households with adolescents reaches 44%. According to educational level, households with incomplete university and incomplete non-university higher education spend the most on average in FAFH, with 247 and 234 Peruvian Soles respectively; households where the head of household has completed secondary school have the highest percentage (31.53%).

The highest expenditure is made by households who obtained a house assigned by the work centre (349.72 Peruvian Soles), followed by households who live in rented housing (254.92 Peruvian Soles). According to the type of housing, households who are buying their house on instalments spend the least (150.38 Peruvian Soles). On the other hand, 66.44% of the population lives in a totally rented house. According to the marital status variable, the households whose head of the household is married spend the most (232.19 Peruvian Soles), and those who are divorced spend the least (125.01 Peruvian Soles). For their part, heads of households who live as cohabitants account for 37.11% of all households with adolescents.

The highest expenditure according to the age of the head of the household is in the group between 16 and 29 years old (250.06 Peruvian Soles), followed by the group between 50 and 59 years old (233.42 Peruvian Soles). On the other hand, the group between 30 and 39 years old spends the least (188.34 Peruvian Soles). The highest expenditure according to the occupation of the head of household is in the group with a dependent job (264.82 Peruvian Soles), which is 48.7 Peruvian Soles more than the average. According to the poverty variable, the extremely poor spend 46.28 Peruvian Soles and non-poor households spend 5.58 times more (258.33 Peruvian Soles). Moreover, households in the urban area spend 65.34 Peruvian Soles more than households living in the rural area. People from the Quechua and Aymara ethnic groups spend the same amount, namely 229 Peruvian Soles (Table 3).

In households with adolescents, when the number of household members increases by one, the probability of spending on FAFH increases by 5%. If the household is located in a rural area, it is 6% less likely to spend on FAFH. The variable native from the Amazon region has 18% less probability of spending on FAFH; similarly, the following variables have less probability of spending on FAFH: Afro-descendant (10%), white (11%), and mestizo (11%), in comparison with those defined as Quechua.

Regarding educational level, when the head of household does not have initial education, he or she is 57% less likely to spend on FAFH consumption compared to heads of households who have no education. Likewise, a head of household with incomplete primary education is 6% more likely to spend on FAFH compared to those with no education. However, having completed non-university higher education has 8% less likely to spend on FAFH, and having completed university higher education has 15% less likely to spend on FAFH compared to those with no education.

Regarding marital status, household heads who are married are 3% less likely to spend on FAFH compared to those who are cohabitants. Moreover, heads of household aged 30 to 39 years, 40 to 49 years, 50 to 59 years, and over 60 years are less likely to spend on FAFH (8%, 6%, 8%, 16%, respectively) compared to those in the 16 to 29 years age group. Heads of households with adolescents who are in dependent work are 5% more likely to purchase FAFH compared to those who have independent jobs.

Households classified as "non-extreme poor" and "non-poor" are 16% and 34% respectively more likely to spend on food consumed outside the home compared to those in extreme poverty (Table 4).

The classification test of the model was 71.91%, sensitivity 94.27% and specificity 21.98%. This explains that, out of every 100 households, the model correctly predicts 71 of them.

The area under the curve was constructed by calculating the probability of prediction according to the result of the Logit model for the probability of purchasing or not purchasing food away from home. The area under the curve of the logit model was 0.691 (0.056-0.977) and the area of those with the probability of spending on FAFH was 0.691 (0-1), var dummy. Moving away from 0.5 discriminates better (see Figure 1).

Figure 1. Roc curve of the Logit model.

Discussion

In recent decades, the consumption of food away from home has become seen as an unhealthy eating behaviour as many of the foods consumed are considered to be not nutritious as they contain harmful substances such as saturated fats, refined sugars and high amounts of sodium, especially in younger populations such as adolescents.²

The expenditure on food consumed away from home for households with adolescents is 216.1 Peruvian Soles and the expenditure for households without adolescents was 56.96 Peruvian Soles less or 15.07 US Dollars. The difference in means is significant.

The description of expenditure on the variable educational level shows large differences, while a head of household can spend 23 Peruvian Soles, the head of the household with a university degree spends 10 times more. Expenditure according to the type of housing (rented, owned and fully paid, owned by encroachment, owned by purchasing in instalments, assigned by work centre, assigned by another household), is above average and does not show large differences. The marital status of the head of household shows divorced people with the lowest expenditure and married people with an expenditure of 107.18 Peruvian Soles more.

There are few studies on models that evaluate the probability of spending on food consumed away from home for households with adolescents, both nationally and internationally. On the other hand, the result of the analysis showed the variables that favour the probability of spending to dependent work, being non-extreme poor and non-poor with a result similar to the work by Lozada-Urbano et al.,⁵⁰ who evaluated a population of households in 2019. In countries like Bangladesh,⁵² street vending is dishonest and there is a lot of adulteration, which discourages wealthy households that come from urban area and with heads of household educated to consume and spend on FAFH.⁵² The prices of foods away from home in the USA were associated with consumption categories of low-income youth.⁵³

In the Lozada-Urbano study in 2019, extended families with more people spent more on FAFH, while in the 2021 study, the number of members in the household favours the probability of expenditure. In addition, incomplete primary education is also in favour of the probability of expenditure; however, the rest of the education variables (initial education, incomplete non-university higher education, complete non-university higher education, complete university higher education), have an expected result for not being favourable. It is likely that education and knowledge about negative factors with food consumption away from home of the heads of household represent a barrier to purchase. Some authors in a study conducted on meals and fats content has shown that the food purchased and consumed in a food court of a shopping mall in Lima was high in fats, even without including calories from beverages and creams.⁵ In countries such as the USA, education has been shown to have a negative effect on obesity.⁵⁴

Some limitations of this study when interpreting the findings for households were that heads of household with incomplete primary education showed a positive probability of expenditure on FAFH. This could be explained that the costs of food result to have a lower value than the cost that includes the cost of food preparation such as household services (electricity, gas and water), household appliances and food storage, also it does not include the cost of time to buy and prepare food.⁵⁵

Variables that do not favour the likelihood of expenditure are being from the rural area, probably in some cases, purchasing food is not allowed since there is no access to shopping malls, and food and fast food businesses. However, a study by Mottaleb in Bangladesh describes that urban households are the least likely to consume and spend on FAFH because of the problem of food adulteration by dishonest vendors.⁵²

None of the racial variables favour the likelihood of expenditure. In the USA, a study with people of African American and Euro-American ethnicity on food selection showed that perceptions of African Americans describe lower cost foods and larger portions than Euro-Americans.⁵⁶

Respondents with higher overall diet quality were more likely to be older and be women with higher educational levels, while they were less likely to be non-Hispanic African descendant and more likely to be of a race or ethnicity other than non-Hispanic white, non-Hispanic African descendant, and Hispanic; and were more likely to have a higher income-poverty ratio.⁵⁵

The ages of the heads of household between 30 to 39 years, 40 to 49 years, 50 to 59 years and being older than 60 years, were not favourable to the probability of expenditure. This result may be due to the fact that it was not worked as a continuous variable or it is explained by the period evaluated corresponding to the second and third wave of the coronavirus disease 2019 (COVID-19) pandemic, where the stay at home was maintained. Some authors consider that it is necessary to consider the form of food delivery during the pandemic. In Peru, there is no evidence on home food delivery.⁵⁷

The results of this study about the poverty variable show the expected results. Regarding the poverty variable (extreme poor, poor and non-poor), and the expenditure on food according to these categories, it is shown that the expenditure is higher in households with adolescents than in those without adolescents, and in the least poor (non-poor). Meanwhile, in households in Australia, they have described some social factors that influence expenditure on foods in the household, such as: the educational level, the household income, the place where they live and the time they have available to purchase unhealthy food.

Likewise, household income is the most predominant factor in the determination of expenditure on food in household, above the socioeconomic level of the household. Meanwhile, the time variable seems to be more related to socioeconomic level with respect to expenditure on food. On the other hand, the comparison between expenditures on food consumed away from and inside the home gave evidence that households with a higher socioeconomic and educational level had less budget expenditure for industrialized and non-industrialized food prepared at home, while for food purchased away from home, mainly in restaurants, more was spent.⁵⁸ In Brazil, the expenditure on food and group of foods grew significantly.⁵⁹

Conclusion

Political efforts are needed so that society assumes and helps to create a food culture both in quantity and quality, regarding the choice of food consumed away from home, due to the high content of critical nutrients such as saturated fats, sodium and sugar, which leads to the increase of a current public health problem such as NCDs. The state, as the governing, executing and regulating entity, should create and prioritize public policies to stimulate food education of the population on nutrition and healthy eating, taking into account the sociodemographic characteristics. These policies should be directed to those who consume and prepare food, as well as the head of household. Regarding academia, it is required to promote research on this topic, which has been so little studied. This way more science and evidence would be developed to propose new nutritional strategies. Also, more promotion should be made in nutrition professionals to develop practical proposals that get concretized in efficient and effective actions in time for the development of counselling and training spaces on issues of food consumption away from home in adolescents.

This work is carried out to contribute to the knowledge of food consumption away from home. Background and studies in other countries, on individuals and households, have already shown that it has a negative effect on households that consume them, increasing overweight and obesity. This result will allow us to know the differences in households and also to compare it with other results that have used the ENAHO national household survey. Likewise, this research will serve as support for future research on the subject matter.

Scientific information on the adolescent population is scarce. Our main contribution is about the variables that favour the probability of purchasing FAFH; the variables used allow us to describe the households that would be consuming more foods that do not help household members to be healthy, especially adolescents.