*Article* **Food Consumption Determinants and Barriers for Healthy Eating at the Workplace—A University Setting †**

**João P. M. Lima 1,2,3,4,\*, Sofia A. Costa <sup>5</sup> , Teresa R. S. Brandão <sup>6</sup> and Ada Rocha 2,3,7**


**Abstract:** Background: A wide variety of social, cultural and economic factors may influence dietary patterns. This work aims to identify the main determinants of food consumption and barriers for healthy eating at the workplace, in a university setting. Methods: A cross-sectional observational study was conducted with 533 participants. Data were obtained through the application of a self-administered questionnaire that included socio-demographic information, food consumption determinants and the main perceived barriers for healthy eating at the workplace. Results: The respondents identified "price" (22.5%), "meal quality" (20.7%), and "location/distance" (16.5%). For women, the determinant "availability of healthy food options" was more important than for men (*p* < 0.001). The food consumption determinants at the workplace most referred to by respondents were related to the nutritional value. Smell, taste, appearance and texture, and good value for money, were also considered important for choosing food at the workplace. Respondents referred to work commitments and lack of time as the main barriers for healthy eating at the workplace. Conclusions: Identification of determinants involved in food consumption, and the barriers for healthy eating, may contribute to a better definition of health promotion initiatives at the workplace aiming to improve nutritional intake.

**Keywords:** food choice; food consumption; university; workplace; determinants; barriers

#### **1. Introduction**

Globalization has caused drastic changes in food patterns within the last decade. These changes resulted in a reduction in the prevalence of malnutrition along with a widespread increase in prevalence of overweight and obesity [1]. An unhealthy lifestyle is one of the major risk factors for chronic diseases in developed countries [2]. Consumer behaviors play a prominent role in the etiology of several chronic non-communicable diseases, including obesity, diabetes mellitus, and cardiovascular diseases, among others, whose prevalence tends to stand still, or even increase [1,3,4].

Sedentary habits and unhealthy eating behaviors are responsible for a significant economic burden through absenteeism and presenteeism [5–8]. Additionally, for employees, unhealthy lifestyle behaviors and obesity might lead to negative effects related to work [9].

**Citation:** Lima, J.P.M.; Costa, S.A.; Brandão, T.R.S.; Rocha, A. Food Consumption Determinants and Barriers for Healthy Eating at the Workplace—A University Setting. *Foods* **2021**, *10*, 695. https:// doi.org/10.3390/foods10040695

Academic Editor: Pascal Schlich

Received: 1 March 2021 Accepted: 22 March 2021 Published: 25 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Research has shown that unhealthy employees and those with an unhealthy lifestyle are less productive at work and have decreased work ability [10–14].

The workplace is recognized as an opportune and fruitful setting for health promotion because of the presence of natural social networks, the possibility of reaching a large number of people, and the amount of time people spend at work [15,16]. Promotion of healthy lifestyles, namely healthy nutritional behavior at the workplace, improves workers' health and productivity [17].

The workplace also offers an interesting context for studying eating behaviors. There is often a high level of consistency in people's working lives, with many workers (particularly those who are office-based, as in this sample) spending most of their time in the same location surrounded by the same group of colleagues [18]. Partly for this reason, a number of eating-related research studies have been conducted at the workplace [19–21].

A wide variety of social, cultural, and economic factors may influence dietary patterns. Intra-individual determinants, such as physiological and psychological factors, acquired food preferences, and knowledge about nutrition can be distinguished from interpersonal or social factors, such as family and partners influence [21].

Food choice determinants are frequently presented in four groups:


In addition to the determinants described above, the individual's psychological state is also assumed as one of the major determinants of the act of eating. Situations of emotional difficulty, states of anxiety and stress, situations of rejection, or loneliness, in more vulnerable individuals, can lead to changes in eating behavior [21].

Several studies concluded that individuals who identified a higher number of barriers for healthier eating habits correspond to those with worse habits [23,24]. The main factors identified by consumers as barriers for healthy eating were lack of time, poor cooking skills, food price, or the lack of healthy choices at food service units [23–26].

Meals eaten at the workplace represent a large contribution to the daily energy intake and influence the balance of the diet [27]. The study "Food and Portuguese Population Lifestyle" [28], identified the factors that influence the food choices of Portuguese adults, and their relationship with socio-demographic and health features [29]. The attribute of "Taste" was the most important factor determining food choice, followed by the "Price" and the "Intention of healthy eating", according to Poínhos et al. [29].

Previous research conducted at different workplaces related to food consumption determinants and perceived barriers, identified that structures and systems within the workplace have a significant role in dietary behaviors. These include the facilities available [30–32], training of staff [33], long hours worked as a result of high workloads and work pressures, and a culture that encourages working through breaks [34,35]. Lack of time for lunch can affect both health and productivity [36,37]. The conflict between promoting a greater range of healthier foods and business constraints has also been previously identified [38].

In order to develop effective workplace interventions for healthy eating, researchers must first consider all the known determinants of eating behavior as potential targets for intervention, such as distinct features of working conditions. In a recent systematic review of factors affecting healthy eating among nurses, the majority of studies found that workplaces often create barriers for healthy eating [20]. Therefore, to define appropriate health promotion initiatives, it is necessary to characterize the determinants involved in food choice, in order to influence food consumption at the workplace. Additionally, to identify perceived barriers for healthier eating habits it is also important for the implementation and assessment of interventions in different scenarios [39,40].

To the best of our knowledge, there are no studies that identify and characterize the determinants involved in food choice in Portugal, especially at the workplace, and it becomes relevant to develop research to better understand this subject. Therefore, this study intends to identify the perceived barriers for healthy eating, and the main determinants of food consumption at the workplace, among university employees.

#### **2. Materials and Methods**

#### *2.1. Study Design and Sample*

A cross-sectional observational study was conducted at a Portuguese university through face-to-face interviews by a trained researcher at the participants' workplace. This university had 3307 employees: 1750 teachers and researchers (academic), 1551 nonteaching staff (non-academic) [41]. A convenience sample was used, stratified by organic units, aiming to represent the study population, allowing researchers to infer conclusions for the study population. Given that the sample corresponds to approximately 15% of the population, it was stratified into teaching and researcher staff, and non-teaching and non-researcher staff; 533 employees were selected. Data collection was performed during labor hours.

#### *2.2. Ethical Issues*

The project was approved by Ethical Commission of the University of Porto, with the number CEFADE 25.2014. The principles of the Helsinki Declaration were respected and the workers under analysis accepted participation in the study through informed consent, after having the purpose and methods involved in the study explained to them individually.

#### *2.3. Questionnaires for Data Collection*

Data were obtained through the application of a self-administered questionnaire. It included socio-demographic information and food consumption determinants at the workplace, and a list of barriers for healthy eating at the workplace. The questionnaire included questions such as the employee's age, gender and marital status. Academic qualifications were also questioned, through a closed answer format composed of nine levels of response (between primary school and PhD or Post-Doc). Employees with academic qualifications higher than bachelor's degree were asked about the training area. Concerning work practices, respondents were asked about the amount of time they spend working at this institution, and the work regime (full-time or part-time). They were asked about the professional category, function performed, with discrimination between teaching and non-teaching activity, and the establishment where they work.

To assess food consumption determinants, a section of the questionnaire was developed through the adaptation of the Food Choice Questionnaire, developed by Steptoe et al. [42] after translation and validation for the Portuguese population by Cardoso and Vale [43]. Steptoe et al. also contributed to the questions of the Food Choice Questionnaire. A Likert Scale of 5 points, from strongly disagree (1) to strongly agree (5) was used in the questions related to determinants. Questions used in the studies "Food and Portuguese Population Lifestyle" and "Food and Portuguese Population Lifestyle" [28,29] were included in the questionnaire. The determinants of the choice of location for lunch in the workplace were also evaluated. Respondents were invited to select the three main factors affecting their choice from a predefined list presented in our results [29,44–47].

The barriers presented to respondents were selected from the literature, and others were added considering individual perceptions of the researchers. Respondents could select as many options from the list as they wanted.

Food offer, quality of meals, prices and food and nutritional intake of employees were analyzed and published in previous research papers [48,49].

#### *2.4. Statistical Analysis*

Data were analyzed using the Statistical Package for Social Sciences version 21.0 ® for Windows. Descriptive analysis was performed, and normality of cardinal variables was tested with Shapiro-Wilk Test. Association between nominal variables was analyzed by chi-square test. Association between ordinals and nominal variables was performed with Kruskal-Wallis tests. Between ordinal variables, or between ordinal and cardinal nonnormal, Spearman correlation was performed. Taking into consideration the differentiation of the sample in terms of age, results were analyzed by age groups, through splitting the sample by the median age (43 years old) to identify younger and older respondents. Cut-off of 0.05 was used as the level of statistical significance. Data were also analyzed according to Multiple Correspondence Analysis (MCA) procedures, which allows for exploring the pattern of relationships of several categorical variables and representing them in few dimensions of homogeneous variables. For this model, sociodemographic variables were included, namely gender, educational level, and professional occupation; lunch setting (lunch brought from home, university food services, restaurants and go home), determinants for the lunch place choice and determinants of food consumption identified from Food Choice Questionnaire [42,43].

#### **3. Results**

#### *3.1. Sample Characterization*

From 533 assessed individuals, 513 were considered valid answers. Participants were aged between 21 and 80 years old (mean 43.3 ± 10.6), mostly females (65.5%) and married (63.4%). About 94% of respondents were full-time workers. Most workers (80.3%) had a university degree and about 35% had a PhD or a Post-Doc diploma. Only 3.3% of respondents did not complete high school education. Of respondents, 34.2% were Teachers, 63.0% were Non-Academic Staff/Researchers and 2.8% had both activities.

The majority of workers had a sedentary activity since 81.5% of them reported spending most of their time seated, and 74.5% characterized their work as not being "very physically demanding".

Only 23.1% of respondents reported following an unhealthy diet at the workplace. Hence, only these workers were asked to point out the barriers for adopting a healthier diet.

#### *3.2. Determinants of Choosing the Place for Having Lunch*

The majority (96.7%) of respondents had lunch every day, however, only 36.1% of them attended the university food service. Of the respondents, 28% had lunch in local restaurants. About 52% of workers brought lunch from home and only 16.2% had lunch at home.

The respondents identified "price" (22.5%), "meal quality" (20.7), "location/distance" (16.5%), "healthy food options" (13.1%) and "lead time" (10.6%) as the most important determinants used to choose the place for having lunch. For women, the option of having "healthy food options" (*p* < 0.001) was more important than for men. Additionally, "location" (*p* < 0.001) and "noise" (*p* = 0.016) were more important for women than for men (Figure 1).

**Figure 1.** Food consumption determinants to choose the place for having lunch per gender. N: Number of individuals

"Price" as a determinant for choosing the place for having lunch was more important in younger respondents (Table 1). This determinant was also more important for those with a lower academic degree (*p* < 0.001) than for those with a higher level of education. Respondents with a higher academic degree referred more frequently to "Location/Distance" of places for having lunch as a determinant of choice. "Meal quality" (*p* = 0.002) and "healthy food options" (*p* = 0.049) were considered determinants for choosing the lunch setting more frequently by teaching staff.


**Table 1.** Food consumption determinants to choose the place for having lunch per age group.

<sup>1</sup> Differences with statistical significance.

Based on results of MCA three main dimensions were identified that explained 33.4% of data variability. The following homogeneous groups of variables were obtained (Figure 2).

**Figure 2.** Food consumption determinants to choose the setting for having lunch (Multiple Correspondence Analysis (MCA) analysis).

#### *3.3. Determinants of Food Consumption at the Workplace*

Determinants of food consumption at the workplace most referred to by respondents (more than 70%) were related to foods rich in vitamins, minerals and fiber, nutritionally balanced, with natural ingredients and no additives, and that contribute to health and weight control. Smell, taste, appearance, texture, and a good value for money were also considered important for choosing food at the workplace.

Based on the results of MCA, two main dimensions were identified that explained 59.9% of data variability, and the following homogeneous groups of variables were obtained (Figure 3).

**Figure 3.** Food consumption determinants at the workplace (MCA analysis).

#### *3.4. Barriers for Healthy Eating at the Workplace*

The participants referred mostly to work commitments and lack of time as barriers for healthy eating at the workplace (Figure 4). From the barriers under analysis, differences between genders were only observed related to knowledge about nutrition. Males identified "Lack of knowledge about nutrition/healthy eating" as a barrier for healthy eating more frequently than women (Table 2). No differences were observed between age groups related to perceived barriers for healthy eating (Table 3).

**Figure 4.** Frequency of perceived barriers for healthy eating at the workplace.



<sup>1</sup> Differences with statistical significance.


**Table 3.** Perceived barriers for healthy eating at the workplace by age group.

In comparing academic with non-academic respondents, significant differences for two distinct barriers were found. It seems that food price is a prohibitive factor for having a healthy diet, essentially for non-academic staff in relation to other individuals (*p* = 0.004). Lack of healthy options for breakfast, lunch and dinner were identified by academic staff more frequently than by non-academics (*p* = 0.012) (Table 4). Concerning other parameters assessed, ranges of age and marital status did not seem to influence the barriers for healthier eating at the workplace.


**Table 4.** Perceived barriers for healthy eating at the workplace by professional occupation.

<sup>1</sup> Differences with statistical significance.

#### **4. Discussion**

Major determinants for choosing a place to have lunch were related to "meal quality", "price", and "location". Working at higher education institutes determines an increased burden of work and responsibilities, most of them extra classes [50], which contributes to work commitments and lack of time to take breaks, prepare, and have healthy meals. Additionally, sensory aspects of food consumption can influence the choice of lunch place. Sensory aspects are usually observed as determinant of food consumption. The cost of meals is more relevant for younger respondents as observed in a previous study [51].

Younger, non-teaching female employees with lower academic qualifications are the group who most frequently bring lunch from home. Bringing food from home is likely associated with higher level cooking skills—more common in the female gender [25]. Additionally, this group also has lower disposable income and hence, bringing food from home allows for more savings.

Lunch location is also determined by other factors. According to other authors, meals outside the home often have a higher energy value and a poorer nutritional profile [27]. Indeed, of the women who bring lunch from home, some do so to ensure a healthier lunch.

On the other hand, teachers with PhD or Post-Doc Diplomas mentioned waiting time as a key decision driver. This is likely associated with a higher level of responsibility, strong focus on work, and consequently, shorter lunch breaks.

In this study, food availability was identified more frequently by academic staff than other respondents. On the other hand, non-academics reported a higher concern, and identified the lack of storage facilities and food preparation areas at the workplace as a barrier. This parallelism on identified barriers could indicate that academics more frequently use university cafeterias, and non-academics bring food from home and use storage and preparation facilities, when available at the workplace, more frequently. These results are in line with the identification of a third barrier, significantly the difference between individuals with different professional occupations. Effectively, non-academics identified the price of healthy food options as a barrier for healthy eating more frequently than academics. Differences in salary between them could explain this result. The perception of these factors could influence the choice of place for having meals—cafeterias, or storage and preparation facilities.

Attending to the wide availability of information about healthy eating, the number of respondents that identify the lack of knowledge about nutrition or healthy eating as a barrier is unexpected. Men identified this barrier more frequently than women. In addition, Yahia observed that men identified the barrier, lack of knowledge about nutrition or healthy eating, more frequently than women, among university students [52].

Universities are a captive environment where staff is restricted to a campus where offices, classes and study facilities are located, and where there is limited choice for food provision [53,54]. The workplace can be a strong determinant of food consumption behavior as it provides convenient access to healthy and/or unhealthy food choices. In a population experiencing time constraints having good food choices at the workplace provides an easy option for refueling [37,48]. Food available at, or near workplaces, is more convenient, low in cost, and sells well [21]. Similar findings were reported by Pinhão et al. in a representative sample of the Portuguese population, where "taste" was the most selected factor, followed by "price" and "trying to eat healthy" [29] as determinants of food choice.

Our results are in accordance with those found by Kjøllesdal in Norwegian adults, showing that people with higher educational levels and in higher income groups ate in staff canteens more frequently than others [55].

According to previous literature, access to healthy foods in the workplace is often limited, compared with an abundance of unhealthy foods present in workplace canteens, onsite shops, and vending machines [46,48,56,57]. According to literature, workers desire a greater variety of healthy and fresh foods compared with the current offerings [46,57–59], which is identified in this research as a barrier for healthy eating. Healthy options also determined workers food choice. Interestingly, some employees felt that food served in the canteen is not balanced with their nutritional needs. The factors that influence food consumption of employees related to healthy options, nutritional value of foods, meal quality, and health and well-being, may be associated with employees' perception of canteen' meals being too high in calories and tailored for physically demanding roles [46].

However, employees also reported that the lunch provided by the work canteen is the only opportunity to have a "proper meal" each day [58]. In the same way, the workplace could be a provider of healthy foods (such as vegetables and fruit) and increase intake of those foods [59,60]. Availability at the workplace is a determinant for food choice and a barrier for healthy eating, the reasons why the availability of facilities where food can be prepared was considered to be an important facilitator of healthy eating [46,59]. On the other hand, the higher cost of healthy options compared with unhealthy options was identified as one of the most significant barriers to healthy eating [46].

The determinants that most influence food choice at the workplace in this study are related to the individual. The identification of knowledge about the health benefits of food is commonly observed, followed by biological determinants such as taste, smell, or the texture of the food, and finally, of an environmental nature related to the quality-price ratio of the food.

Food choices of men, with higher academic qualifications and belonging to the teaching staff, are determined by food taste and texture, and by availability and price-quality relationship. Additionally, they value the potential benefits of food, and their food choice is determined by them. The influence that foods can have on well-being is also important, such as choosing foods that help maintain alertness and support emotional health.

Regardless of gender, among professors with higher academic qualifications, food choice is determined by cultural, religious or ethnic beliefs, political ideologies, the clarity and environmental responsibility of packaging products, and medical advice regarding the intake of certain foods. On the other hand, among individuals with lower academic qualifications, these determinants have a reduced importance.

In fact, food choice is a complex result of preferences for sensory characteristics, combined with the influence of non-sensory factors, including food-related expectations and attitudes, health claims, price, ethical concerns and mood, as already reported by other authors [47,61]. Regarding these concerns, the availability of healthy food options at the workplace, namely in cafeterias, is very important. On the other hand, the inability to prepare meals was also identified as a barrier for healthy eating, pointing to a need to improve cooking skills, for example, by the inclusion of this topic in the school curriculum.

Only a small proportion of respondents perceived barriers for adoption of a healthy diet. Other authors observed similar results [25,26,62]. Healthier environments should be promoted to facilitate healthy eating and fighting chronic diseases such as obesity [63]. However, of all variables tested, only the price and lack of knowledge about nutrition/healthy eating showed significant differences between respondents. Some studies

have shown that people that identify a higher number of barriers are those that follow unhealthy eating habits more frequently [24,63].

The barriers identified in this research are related only to individuals that are considered as having unhealthy eating at the workplace. Future works should also include those who are considered as having healthy habits.

Strategies to promote healthier food habits aim at reducing barriers to access healthy options and increasing opportunities for employees to make healthier food choices. Implementation includes provision of healthier options, improved accessibility, and establishment of mandatory policies to provide healthy options or restrict less healthy offerings at the workplace [16].

Some limitations were identified in this study. Lack of information concerning income that impair conclusions potentially explained by this. Another limitation was related to the usage of different tools to access food determinants for choosing the place to have lunch, and the determinants of food consumption in general. However, the fact that the tool used to access the determinants for choosing the place to have lunch was used in another Portuguese study with a national representative sample, motivates the researchers to that procedure. The use of a convenience sample determined a higher proportion of non-academic staff as they were more available for data collection.

#### **5. Conclusions**

The most important determinants identified by respondents choosing the place for having meals were "meal quality", "price", and "location/distance". For women, the availability of "healthy food options" was more important than for men.

Our results seem to demonstrate that gender, marital status, academic degree and main professional occupation, are related to the choice of the place for having lunch. Differences were found between gender, marital status and age ranges, in terms of factors-affecting food choice at the workplace. A higher concern with nutritional value of food was observed for younger respondents, individuals living alone, and women.

Gender and academic degree are relevant in food choice. Factors influencing individuals with a low academic degree were previous food habits, price, and quality of meals, in determining the choice of place for having lunch at restaurants or at home. On the other hand, women with a high academic degree prefer to bring meals from home as they find them healthier.

Related to determinants of food choice in general, MCA analysis reported the major differences related to academic degree and main occupation, with lower academic degree individuals being not influenced by external determinants, since their food choice was mainly influenced by previous food habits. Higher academic degree employees in general are influenced by nutritional value of food and its relationship to health and well-being, packaging, and health professional advice, the reason why strategies to promote healthy eating in these scenarios are necessarily different. If we could design a healthy eating program based on information about the nutrition value of food and health, namely through packaging, our results would show clearly that this option could be adequate for teachers and other employees with high academic degrees, but not for others that probably need personal counseling to change previous food habits.

This work also identified lack of time, work commitments, and lack of healthy options for having meals at the workplace as barriers for healthy eating. Educational level, professional occupation, and gender were the socio-economic characteristics evaluated that influenced the perception of barriers for healthy eating.

These results may contribute to a better definition of strategies to promote healthy eating in these scenarios and show that different strategies are needed for different target groups to reduce barriers once they are perceived differently by individuals.

**Author Contributions:** Conceptualization, J.P.M.L. and A.R.; formal analysis, S.A.C. and T.R.S.B.; investigation, J.P.M.L. and A.R.; writing—original draft preparation, J.P.M.L.; writing—review and editing, A.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by national funds through FCT—Foundation for Science and Technology within the scope of UIDB/05748/2020 and UIDP/05748/2020.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethical Commission of the University of Porto (protocol code CEFADE 25.2014 at 22/10/2014).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The work was a part of João Lima's doctoral thesis.

**Acknowledgments:** Authors thank Graça Neto, English Professional, for English grammar and structure revision of the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Demographic, Anthropometric and Food Behavior Data towards Healthy Eating in Romania**

**Anca Bacârea 1, Vladimir Constantin Bacârea 2,\*, Cristina Cînpeanu 3, Claudiu Teodorescu 3, Ana Gabriela Seni 3, Raquel P. F. Guiné <sup>4</sup> and Monica Tarcea <sup>3</sup>**


**Abstract:** Background: Each country has specific social, cultural, and economic characteristics regarding the motivations for improving health. The aim of this study was to evaluate demographic characteristics, anthropometric data, and elements related to food behavior and health, as well as Romanians' motivations towards healthy eating. Methods: This is a descriptive cross-sectional questionnaire based study enrolling 751 Romanian participants, which was carried out in in 2017– 2018. Results: We obtained a positive correlation between age and Body Mass Index, and this was maintained also when we analyzed the two genders separately, being, however, even stronger for women. The number of hours/day spent watching TV or in front of the computer was positively correlated with both age and BMI. In general, with aging, there is an increasing concern regarding the practice of a healthy diet. The higher education level was significantly associated with healthier choices. Conclusions: The study of the three dietary dimensions, food properties, health attitudes, and dietary behavior, vis-à-vis various disorders revealed that the group most concerned of their diet was those who suffered from cardiovascular disorders.

**Keywords:** health; motivation; BMI; food behavior; education

#### **1. Introduction**

There have been various studies regarding the impact of social and cultural factors upon different communities' food behavior [1]. There is substantial evidence that social norms regarding food consumption strongly effect food choice, quality, and quantity consumed [2].

The globalization of agrifood systems has increased the availability and variety of foods through in food production and distribution changes. On one side, agricultural priorities rely on production and processing systems, markets, and livelihoods, with more concern for food safety and less care about general public health issues. Conversely, traditional public health focuses on agricultural issues that affect food security and the potential role of agriculture in preventing food-related diseases. We need to consider multidisciplinary aspects and the complex relationship between agribusiness, food consumption patterns, and health [3].

Adopting healthy diets can improve the nutritional behaviors and the status of population health. The guidelines released by the World Health Organization (WHO) establish a substantial reduction in the consumption of dairy products (by 28%), animal fats (by 30%), meat (13%), and sugar (by 24%) and a substantial increase in the consumption of cereals

**Citation:** Bacârea, A.; Bacârea, V.C.; Cînpeanu, C.; Teodorescu, C.; Seni, A.G.; Guiné, R.P.F.; Tarcea, M. Demographic, Anthropometric and Food Behavior Data towards Healthy Eating in Romania. *Foods* **2021**, *10*, 487. https://doi.org/10.3390/ foods10030487

Academic Editor: Mari Sandell

Received: 14 January 2021 Accepted: 19 February 2021 Published: 24 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

(by 31%), fruit (by 25%), and vegetables (21%) in order to reduce the burden of chronic diseases [4,5].

In the scientific literature, many variables have been used to analyze consumer behavior in the food market, the most frequent being the socio-demographic factors, motivations and attitudes, religious traits, and cultural and social background, along with geographical variability and lifestyle [6–8].

Over time, the Romanian people have undergone various lifestyle changes by adapting to food availability, information sources, and social status. In the early 1990s, the American model was adopted at the same time as the country's modernization. Fast-food products emerged, and the number of supermarkets increased [9]. Highly processed foods with calorie-dense content, rich in carbohydrates, lipids, flavor enhancers, and food additives, have become popular among children and adolescents. Their popularity continues to grow insidiously due to the lack of the population's targeted education [10]. Subsequently to the increase in obesity and the frequency of cardiovascular diseases, based on the development of nutrition-related mass-media projects, healthy food has been promoted in the last years. Consumers are urged to adopt a balanced diet to prevent food behavior disorders [11,12]. Many studies have focused on the importance of proper labelling ("low fat", "low sugar", "special price") to support consumers' choices [13,14]. A study shows that the probability of buying a product is higher when the price is low and the product is perceived healthier or tastier [15]. Another study found that price discounts seem to have ambiguous effects by promoting the purchase of healthy products, but also leading to increased calorie-dense purchases [16]. Unfortunately, most consumers are not aware or interested in reading food labels, depending on the social and cultural profile [17]. Many questions on what people understand about healthy eating still need to be answered.

Advertising is one of the major factors influencing the purchasing behavior of the population. Most of the TV advertising spots on food are dedicated to sweet, salty, and fat products (approximately 89%) [18]. Besides nutrients, people also check for food composition. An important role in choosing food products is played by the neo cortex, emotional eating, and education as well [19].

Worldwide, food behaviors are linked with the risk of occurrence of obesity [20], cardiovascular diseases [21], diabetes [22], respiratory disorders [23], psychiatric disorders [24], and cancer [25]. The role of maternal obesity on foetal development, birth outcomes, and child health is also recognized [26].

Romania, like many other countries, is making efforts to promote healthy eating. In order to do that, some particular aspects determining the motivations with regard to healthy eating should be considered and generated: like socio-demographic, cultural, economic, emotional, and environmental factors [27]. Despite the benefits of healthy eating, many people still prefer unhealthy food, and this indicates the need for more efficient community interventions.

The aim of this study was to evaluate demographic data, anthropometric data, and elements related to food behavior, as well as the Romanian people's motivations towards healthy eating, as the first step for further development of health policies and strategies to improve nutritional behavior. To the best of our knowledge, this is the first Romanian study of its kind, and it is a part of a multinational project entitled "Psycho-social motivations associated with food choices and eating practices (EATMOT)" comprising 16 countries (Argentina, Brazil, Croatia, Cyprus, Egypt, Greece, Hungary, Italy, Latvia, Lithuania, Macedonia, Netherlands, Poland, Romania, Serbia, Slovenia, and The Unites States of America).

This is a pilot study that aims to emphasize the importance of targeted education and community intervention, based on Romanian culture, attitudes, and motivations regarding healthy diets, and also aims to be a starting point for nutritional programs to be developed for Romanians all over the world.

#### **2. Materials and Methods**

This is a descriptive cross-sectional questionnaire-based study, targeted to evaluate the Romanian people's motivations towards healthy eating, carried out during October 2017 and March 2018. The questionnaire was developed and validated within the EATMOT project by Ferrão et al. [6], and then it was translated into Romanian language. Our study includes 751 Romanian participants, from different regions of the country; thus, the study is country representative. The study was approved by the Ethics Committee of the University of Medicine and Pharmacy, Science, and Technology "G.E. Palade" Targu Mures and was conducted in accordance with the Helsinki Declaration.

We included only adult people, aged 18 and older, who fully completed the whole questionnaire. Participants had to answer two sets of questions: the first set of questions referred to demographic data, anthropometric data, and elements related to behavior and health, and the second set referred to motivations for health. The analyzed parameters in the first set of questions were age (we divided the population into five age categories: 18–29, 30–39, 40–49, 50–59, and ≥60 years old), gender, weight, height, environment (urban, suburban, rural), the last level of studies completed (general school, high school, college), marital status (single, married/living together, divorced/separated, widow), current employee status (employed, unemployed, retired, working student), field of activity/specialization in certain areas (nutrition, food, agriculture, sports, psychology, health-related activities), responsibility for eating, physical activity (never, sporadic, occasionally, moderate, intense), hours/day spent in front of the TV or computer, own opinion about having a healthy/balanced diet (never, rarely, sometimes, frequently, always), and the presence of chronic disorders (cardiovascular disease, diabetes mellitus, high cholesterol, high blood pressure, gastric disorders, intestinal disorders, obesity, or others). We calculated the BMI in kilograms divided by the square of height in meters (kg/m2) based on the declared weight and height. We looked for significant associations and differences among age, BMI, and the studied parameters.

The second set of questions comprised ten items, as follows:

Q1—I am very concerned about the hygiene and safety of the food I eat

Q2—It is important for me that my diet is low in fat

Q3—Usually, I follow a healthy and balanced diet

Q4—It is important for me that my daily diet contains a lot of vitamins and minerals

Q5—I do not avoid foods, even if they may raise my cholesterol

Q6—I try to eat foods that do not contain additives

Q7—I do not eat processed foods, because of their lower nutritional quality

Q8—It is important for me to eat food that keeps me healthy

Q9—I do not avoid foods, even if they may raise my blood glycaemia

Q10—I avoid foods with genetically modified organisms

The items included in the second set to answer all ten questions regarding health motivations were: totally disagree, disagree, neither agree nor disagree, agree, or totally agree. The ten questions investigated the participants' interest in food composition (Q1, Q6, and Q10) and healthy properties (Q2, Q3, Q4, Q5, Q7, Q8, and Q9), and together they aimed to show the general picture regarding Romanians' motivations for a healthy diet. Q3 is a frequency-based question, whereas the others are attitudinal questions. Q5 and Q9 were reversed, the question referring to the healthiest attitude being on the left (disagree), compared to the other questions, where the healthiest attitude was on the right (agree). Because these items were measuring different things, we opted to create two composite scales, one for food properties (Q1, Q6, and Q10) and one for health attitudes and motivations (Q2, Q4, Q5, Q7, Q8, and Q9), leaving Q3, referring to healthy diet frequency, to stand alone. We decided to group the questions in this manner, based on face validity and based on the results of the previous studies [28]. In order to do the statistical analyses, we first reversed Q5 and Q9.

We used Microsoft Excel for Mac 2011 (Microsoft Corporation, Redmond, WA, the USA) for data collection and handling, and Graph Pad Prism demo version (Graph Pad Software, La Jolla, CA, the USA) and Epi Info version 7 (Centers for Disease Control and Prevention, Atlanta, GA, USA) and SPSS Statistics v.25 for statistical analyses. We used statistical methods to provide mean and SD for continuous variables or median and range for discrete variables, and absolute and relative frequency counts for categorical variables. The Student *T*-test and Mann–Whitney U test were used as appropriate statistical tests to compare continuous variables between the groups (normal or non-Gaussian distribution); for correlations, we used the Pearson or the Spearman test according to variables distribution. To establish a mean difference between several continuous variables, we used the ANOVA test for Gaussian distributions and the Kruskal–Wallis test for non-Gaussian distributions [29]. A *p*-value under 0.05 was considered statistically significant. The item analysis was performed using the Pearson correlation coefficients, and the associations were interpreted as not existing (r = 0), very weak (0.00<r< 0.10), weak (0.10 ≤ r < 0.30), moderate (0.30 ≤ r < 0.50), strong (0.50 ≤ r < 0.70), very strong (0.70 ≤ r < 1), or perfect (r = 1), according to the value of r [6]. The internal consistency of the scales was evaluated by using Cronbach's alpha, according to Marôco [30], as follows: over 0.9: excellent, 0.8–0.9: very good, 0.7–0.8: good, 0.6–0.7: medium, 0.5–0.6: reasonable, below 0.5: bad.

#### **3. Results**

The characteristics of the studied population: socio-demographic data, environment, professional areas, physical activity, and medical history can be found in Table 1.


**Table 1.** Demographic, anthropometric data, and elements related to food behavior and health status of the studied population.


#### **Table 1.** *Cont.*

(1) N = number of participants.

The BMI values were calculated for the whole sample and varied between 15.05 and 43.57 kg/m2, being on average 24.59 ± 4.34 kg/m2. In Table 2, we evaluated our subjects' BMI and age in relation to the studied parameters.

We obtained a positive correlation between age and BMI, and this was also maintained when we analyzed the two genders separately, this correlation being stronger in women. When analyzing the marital status, we obtained statistically significant differences between single vs. married (*p* < 0.0001) and between single vs. divorced (*p* = 0.0382), but not between single vs. widow (*p* = 0.2386). A continuation of this study to include a higher number of subjects in the widowed category is needed to confirm this result. This was also the case of agricultural worker as a subcategory of professional activity, for which we obtained significantly higher BMI, but the number of cases for this category was lower.

Regarding the number of hours/day spent r in front of the TV or computer, we obtained positive correlations for both age and BMI.


**Table 2. The** relationship among BMI, age, and studied parameters.


**Table 2.** *Cont.*


**Table 2.** *Cont.*

(1) N = number of participants; (2) BMI = body mass index; (3) *p* < 0.05 is considered significant; (4) Spearman test; (5) Spearman correlation; (6) Pearson test; (7) r = Pearson correlation; (8) NA = not applicable; this group was excluded from the analysis due to low number of cases; (9) *T*-test; (10) Kruskal–Wallis test; (11) ANOVA test; (12) Mann–Whitney test; rx = correlation between hours/day spent watching TV or in front of the computer and age; ry = correlation between hours/day spent watching TV or in front of the computer and BMI; (12) N.P. = not performed, because it is not relevant for the current research to compare BMI according to different pathologies (obesity is included, and there are subjects with more than one of the conditions asked).

We obtained no significant associations among BMI, environment, current professional activity, responsibility for eating, and physical activity.

We displayed in Table 3 the results of participants' answers to questions regarding motivations towards healthy eating.


**Table 3.** Results of options regarding the motivations for health.

1—totally disagree, 2—disagree, 3—neither agree nor disagree, 4—agree, 5—totally agree.

Except for Q2, where the highest percentage was obtained for the item "neither agree nor disagree", for the other questions, the highest percentage was registered for items "agree" and "totally agree".

Table 4 shows item–item correlations for the group of questions investigating participants' attitudes toward food properties. The results indicate moderate correlations. The value of Cronbach alpha was 0.689, which is a medium value, based on which we accepted all three questions in composite scale for food properties.


**Table 4.** Item–item correlations for the composite scale investigating food properties (1).

(1) Cronbach alpha = 0.689, \*\* Correlation is significant at the 0.01 level (2-tailed).

Table 5 shows item-item correlations for the group of questions investigating health attitudes and motivations. Because of the negative, very weak, and weak associations between the reversed Q5 and Q2, Q4, and Q7, between Q7 and Q4 and Q5R, and between the reversed Q9 and Q2, Q4, Q5R, Q7, and Q8, we considered eliminating Q5, Q7, and Q9. Q5 and Q9 are negative questions, and the answers were probably inconsistent due to the participants' lack of attention.

**Table 5.** Item–item correlations for the composite scale investigating health status and motivations (1).


(1) Cronbach alpha = 0.517; (2) Q5R = reversed Q5; (3) O9R = reversed Q9, \*\* Correlation is significant at the 0.01 level (2-tailed).

When eliminating these three questions, the Cronbach alpha increased to 0.712, which was interpreted as a good value.

The associations between composite scale for food properties, refined scale for health attitudes, reported dietary behavior (Q3), and investigated variables (*p* values) are shown in Table 6.

With aging, there was an increasing concern for a healthy and balanced diet, but this was not reflected when analyzing food properties in relation to age, perhaps because of lack of knowledge, especially for Q10. However, there was a positive correlation between aging and the three studied dimensions (for food properties and aging r = 0.1408, *p* = 0.0001; for health attitudes and aging r = 0.1643, *p* = 0.0001; for dietary behavior and aging r = 0.1490, *p* = 0.0001). The age category of 18–29 was the most interested in food properties, although this was statistically not significant. The age category ≥ 60 was the most concerned age group about their health and a healthy and balanced diet compared to the other four age groups. The low number of subjects included in the age category ≥ 60 (N = 44) was a limitation of our study, and our results should be confirmed by including a larger population in the study. Women answered more with "agree" and "totally agree" to health attitudes and dietary behavior (Q3) questions than men. Regarding food properties, there was no difference between men and women.

We obtained significantly more "agree" and "totally agree" answers for item Q3 in the case of participants who came from urban environments. For composite scale investigating food properties and health attitudes, respondents from urban environments answered more with "agree" and "totally agree", even if this was statistically not significant.

College education level was significantly associated with health motivations for items investigating health attitudes (college education 52.84% vs. high school 39.52%) and dietary behavior (college education 48.02% vs. high school 31.14%). Even if this was statistically not significant, persons with a college education were more preoccupied with food properties than the persons belonging to the other educational groups.

There were no correlations between BMI, food properties, and attitudes toward health. For Q3, there was a weak negative correlation with BMI.


**Table 6.** Associations between food properties, health attitudes, and reported dietary behavior related to the investigated variables (*p*-value (1)).

(1) *p* < 0.05 was considered significant; (2) Chi square test fornxm table; (3) Spearman test.

Most of the respondents practicing occasional and moderate physical activity answered agree or totally agree for all three investigated dimensions.

TV/computer hours were not correlated with food properties, healthy attitudes (very weak correlation r = 0.0901), or dietary behavior.

People with cardiovascular disorders were more often preoccupied with a healthy diet (92.59% of those who reported cardiovascular disorders answered "agree" and "totally agree" to the questions investigating health attitude). Although statistically not significant, people with cardiovascular disorders answered more often "agree" and "totally agree" for both composite scales: food properties and dietary behavior (Q3). Of those who reported cardiovascular diseases, 7.41% also reported hypercholesterolemia.

People reporting gastric disorders were preoccupied with all three studied dimensions, and those having intestinal disorders were especially concerned about food properties. This finding is not maintained for composite scales investigating health attitudes and dietary behavior, where the individuals without intestinal disorders are more concerned about healthier choices, although statistically not significant.

Persons who reported obesity show a lack of interest regarding all three investigated dimensions, which was significant for reported dietary behavior and not significant for food properties or heath attitudes. We did not interpret the category including other disorders because of the low number of subjects in this group.

We found a significant relationship (*p* = 0.0000) between reported dietary behavior and health attitudes: 224 subjects answered "agree" and 116 subjects answered "totally agree" for both composite scale investigating health attitudes and for Q3, investigating dietary behavior. Analyzing the different age groups, we obtained the following percentages, showing the concordance in responses "agree" and "totally agree" for these two dimensions: 36.53% for the age category 18–29 years old, 46.87% for the age category 30–39 years old, 50.80% for the age category 40–49 years old, 48.48% for the age category 50–59 years old, and 59.09% for the age category ≥60 years old. In other words, with aging, participants' answers are consistent with healthier choices.

#### **4. Discussion**

Modern lifestyle induces harmful behavior regarding eating and physical activity [31]. Altered behaviors are a growing problem in Romania, such as in other countries. Obesity is one important disease associated with unhealthy eating behaviors, and the fact that Romania is a middle-income country can contribute to the obesity epidemic spreading [32]. As our results indicated, obesity was the most frequent health problem reported by the participants in this study (6.66%). Since we have already shown a significant positive association between BMI and glycaemia in the age category older than 22 [33], we consider it appropriate to evaluate the BMI in relation to demographic, anthropometric data, and elements related to behavior and health. Similar to the results of Abdella et al. [34], age positively correlated to BMI. We also demonstrated a positive correlation among BMI, age and the number of hours spent in front of the TV or computer. In their study, Martínez-Moyá et al. [35] proved that the number of hours spent watching television and lower physical activity were significantly associated with a higher BMI in young adults [35]. Other studies showed that an increased screen time spent was significantly associated with the risk of obesity, but not the physical activity level [36], and watching television is the leading sedentary activity in association with obesity [37]. This relation was also found in children; according to Golshevsky et al. [38], higher BMI was associated with more hours spent watching television, and less time spent in organized sports activities. The increased number of hours spent in front of the TV with aging can be related to life cycle changes. When we separately analyzed the relationships between the two genders, the correlation was stronger for women, possibly because of the hormonal changes related to aging. When comparing the BMI between the two genders, women had lower mean BMI values than men (23.97 vs. 25.91), possibly because women's mean age was lower than men's (36.6 vs. 41.04).

Many studies emphasized the role of education in BMI control [39,40]. However, we obtained higher BMI values for the group with college education compared to the group with high school. Some explanations for this finding could be easy access to college education and the lack of physical activity. On the other hand, the group with a college education has the highest mean age, so the BMI can be attributed to aging. Strategies to improve knowledge about healthy eating must be developed to have a better weight control and focus on different age categories.

There was no correlation between the environment and BMI, and this is quite normal considering the people's migration and easy access to information regardless of the native environment.

The marital status also influenced the BMI, married persons having higher BMI values compared to single persons. This is in concordance with the results of other studies [41,42]. It is unclear how marital status affects BMI, probably by changing the body weight-related perceptions and eating behaviors [42]. Some studies indicated that increased BMI affects the status of the employee because of health problems or because it decreases the chances to be employed [43,44]. We did not find BMI differences between employed and unemployed participants, probably because in Romania many young adults prefer not to work because of low income, and they benefit from social assistance (33.41 vs. 41.39 years).

Normally, professional activity is associated with different knowledge regarding health and with different physical activity level. The highest mean BMI value was obtained for those working in agriculture, but the result is debatable due to the small number of subjects in this group.

Even if we did not find a statistical difference for BMI between the different categories of physical activity, we must notice that the mean age of subjects performing intense and moderate physical activity is lower than in those with no or sporadic physical activity. This is a good aspect, showing the concern of younger adults for their health. Many studies showed the benefits of physical activity upon health, associated with a decreased risk of cardiovascular events [45] and better control of blood glucose level [46]. Carraça et al. [47], in their recent study, identified a behavioral pattern showing that adults who are not interested in physical activity are women, have a higher BMI, have been less educated, and are unemployed. Their eating habits are more likely to be less healthy, and they perceive more barriers when it comes to physical activity [47]. Another study shows that barriers to healthy eating and/or physical activity significantly influenced BMI, the level of physical activity, stress, and fruit and vegetable intake [48].

Consumers' beliefs and knowledge about healthy foods are variable. A food is considered healthy in general if it is low in total fats and saturated fats, and meets certain requirements regarding cholesterol and certain vitamins or minerals content [2,49,50]. In his study, Lusk identifies four categories on how healthy food should be defined: based on food nutrients, the entire composition of the food, nutrients from the whole diet, and based on holistic consumption patterns, and the respondents were almost equally distributed among these categories [28].

The fact that a person has adequate knowledge and answers the questions accordingly, does not always mean that this knowledge is applied in everyday life. In a study, the participants showed adequate nutrition knowledge, but eating behavior was strongly influenced by social and physical environmental factors [51]. Mete et al. [52] underline the role of social media in improving healthy food choices by promoting healthy eating information.

This is the motivation for our analysis of the questions in three dimensions—food properties, health attitudes, and separately the general question for dietary behavior (Q3). Looking at food properties (hygiene, additives, and genetically modified organisms) in relationship with the studied parameters, we can see that more physically active people and those having gastric and intestinal disorders were more preoccupied with these parameters. The associations were also statistically significant between physical activity, health attitudes, and dietary behavior (Q3), indicating that these people were concerned and motivated to maintain their health. Genetically modified food is a controversial subject and involves important knowledge [53]. People with gastric and intestinal disorders relate food properties to their disease. We assume that this is why they were more preoccupied with healthy diets compared with other people suffering from other pathologies.

With aging, people were more preoccupied with making healthier choices (significant associations between age and refined composite scales regarding health attitudes and Q3), probably because of changes related to aging and occurring pathologies. This is similar to the results of Whitelock et al. [54], where the participants described efforts focusing on avoiding foods high in fats and sugar content. More women than men had perceptions compliant with a healthy diet, possibly because they have better nutritional knowledge and interest for it [27,55,56].

We obtained significantly more answers compliant with a healthy and balanced diet in general for urban areas. However, this difference was not maintained concerning food properties and health attitudes. This can be due to an increased general interest in a healthy diet that is not translated into specific choices. Education level is essential in connection with BMI and various metabolic diseases [40,57]. As expected, we obtained more correct answers in the college group regarding healthy diet and fat, vitamins, and minerals content, but these answers were not reflected in the case of a lower BMI.

In one study about the perception of healthy eating in Romania, it was shown that tradition is very much related to eating behavior, and was correlated with BMI [31]. Lotrean et al. [58] performed a study among Romanian students and revealed three main dietary structures: two of them protective against becoming overweight, but different regarding physical activity, and the third one (fast food diet) associated with higher BMI and lack of daily physical activity [58]. We believe that people have somewhat imbalanced attitudes about food and healthy eating, which could significantly affect the transposition of beliefs, knowledge about healthy eating, and attitudes into behavior. The study subjects having different disorders showed lack of interest regarding healthy eating choices, which can be a contributing key factor to the evolution of their diseases.

According to Nagata et al., after a seven-year follow-up of young adults with overweight/obesity and unhealthy weight control behaviors at baseline, they still had higher BMI than those without unhealthy weight control behaviors [59].

We found a significant association between health attitudes and cardiovascular disorders, which is logical. People with cardiovascular diseases are more preoccupied to have a particularly low-fat diet.

Traditionally, recommendations were made for individual nutrients consumption such as saturated fats, sugar, sodium, and cholesterol in the diet, because they are usually over-consumed by many people and are linked to the development of chronic diseases. These can also lead to erroneous effects [60–62]. One controversial topic was the association between saturated fats and cardiovascular diseases without considering substitute nutrients and cholesterol, and another one is the association between cholesterol and cardiovascular diseases, which is confused with the intake of saturated fats [63]. The contemporary dietary guide recommends healthy dietary models, with an emphasis on food-based recommendations. The diet as a whole, meaning the combinations and the amounts of food (nutrients) we eat daily, is an essential determinant of health [63].

Our study has some limitations because of gender differences (more women than men) and low number of participants in some categories (e.g., agriculture as a field of activity, age group ≥ 60 years old). Additionally, the participants were not from all counties of the country, so the study is not 100% country representative. We calculated the BMI based on the self-reported values for weight and height, so some bias occurred.

We showed that with aging, people were more preoccupied with making healthier choices. However, it is not only the occurrence of various diseases that should make people aware of healthier choices. Hence, there is a need for an intensive national strategy for health motivations. According to age groups, this strategy should be addressed differently, knowing that radical changes in lifestyle are difficult to accept with age. Another important aspect is the translation of information about healthy choices into real choices. According to our findings, obesity was the most frequent health problem reported by the participants in this study, and despite our expectations, we obtained higher BMI values for the college education group, although they chose the correct answers for their health when asked. This indicates the need for a long-term strategy to motivate people to make healthy eating choices, starting with children and involving also their families [58].

#### **5. Conclusions**

We obtained a positive correlation between demographic parameters and the BMI in the Romanian population; also their healthy food behaviors were stronger for women. The number of hours/day spent watching TV or in front of the computer was positively correlated with age and BMI. The higher education level was significantly associated with healthier choices regarding nutrition practices and motivations. Regarding the associations between the sociodemographic characteristics and different disorders, we observed that the subjects with cardiovascular disorders were more preoccupied with healthier diets in most cases.

Nutritionists, specialists in medicine, and food stakeholders should promote healthy diets through adequate sources of information aimed at target groups. They should develop a more efficient strategy to motivate people to make healthy eating choices and improve Romanian food behavior.

**Author Contributions:** Conceptualization: A.B. and R.P.F.G.; methodology: A.B. and V.C.B.; software: V.C.B.; validation: A.B., V.C.B., and R.P.F.G.; formal analysis: A.B. and V.C.B.; investigation: A.B., V.C.B., M.T., C.C., C.T., A.G.S., and R.P.F.G.; resources: R.P.F.G.; data curation: A.B. and V.C.B.; writing—original draft preparation: A.B.; writing—review and editing: A.B., V.C.B., M.T., and R.P.F.G.; visualization: A.B. supervision: A.B. and R.P.F.G.; project administration: R.P.F.G.; funding acquisition: R.P.F.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Research Center CI&DETS (Polytechnic Institute of Viseu, Portugal) with grant n.º PROJ/CI&DETS/CGD/0012. The APC was funded by FCT-Foundation for Science and Technology (Portugal), scholarship number UIDB/00681/2020.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Data are available from the corresponding author upon reasonable request.

**Acknowledgments:** This work was prepared in the ambit of the multinational project EATMOT from CI&DETS Research Centre (IPV—Viseu, Portugal) with reference PROJ/CI&DETS/CGD/0012. This work is supported by Portuguese National Funds through the FCT—Foundation for Science and Technology, I.P., within the scope of the project Refª UIDB/00681/2020. Furthermore, we would like to thank the CERNAS Research Centre and the Polytechnic Institute of Viseu for their support.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Food Choice Determinants and Perceptions of a Healthy Diet among Italian Consumers**

**Rungsaran Wongprawmas <sup>1</sup> , Cristina Mora 1,\*, Nicoletta Pellegrini <sup>2</sup> , Raquel P. F. Guiné <sup>3</sup> , Eleonora Carini 1, Giovanni Sogari <sup>1</sup> and Elena Vittadini <sup>4</sup>**


**Abstract:** Healthy food choices are crucial for a healthy lifestyle. However, food choices are complex and affected by various factors. Understanding the determinant factors affecting food choices could aid policy-makers in designing better strategies to promote healthy food choices in the general public. This study aims to evaluate the food choice motivations and to segment consumer groups, according to their food choice motivations, in a sample of 531 Italian consumers (collected by convenience sampling), through offline and online survey platforms. K-means cluster analysis was applied to identify consumer groups using six food choice motivation categories (health, emotional, economic and availability, social and cultural, environmental and political, and marketing and commercial). The results suggest that the strongest determinants for the food choices of Italian consumers are Environmental factors and Health. Two consumer profiles were identified through the segmentation analysis: Emotional eating and Health-driven consumers. The respondents were found to have a good awareness of what comprises a healthy diet. There is a potential market for healthy and sustainable food products, especially products with minimal or environmentally friendly packages. Food labels and information strategies could be promoted as tools to assist consumers to make healthy food choices.

**Keywords:** food choices; eating determinants; healthy diet; emotions

#### **1. Introduction**

The food that we consume affects our future health. Diet-related non-communicable diseases (NCDs), such as obesity, type 2 diabetes, cardiovascular disease, hypertension, stroke, and some types of cancer, have been increasingly causing health problems in both developing and developed countries [1,2]. Policy-makers have been trying to introduce several different tools to encourage populations to consume healthier foods and reduce their intake of unhealthy foods, through initiatives such as nutritional education programs and fiscal programs (i.e., sugar drink taxes), among others. Despite these attempts, obesity has greatly risen in the past two decades, even in countries where the rates have been historically low, such as Italy [3].

In Italy, obesity among adults increased from 9% in 2003 to 11% in 2017. Although obesity in adult remained below than the EU average (15%), nearly one in five 15-year-olds in Italy (18%) were overweight or obese in 2013–2014, a share close to the EU average [4]. This raised public policy concern, as excess weight among children and adolescents could affect the population's health in the long run. In order to design appropriate policy tools to increase healthy eating, the motivation behind food choices should be understood and defined.

**Citation:** Wongprawmas, R.; Mora, C.; Pellegrini, N.; Guiné, R.P.F.; Carini, E.; Sogari, G.; Vittadini, E. Food Choice Determinants and Perceptions of a Healthy Diet among Italian Consumers. *Foods* **2021**, *10*, 318. https://doi.org/10.3390/foods10020318

Academic Editor: Koushik Adhikari Received: 30 December 2020 Accepted: 30 January 2021 Published: 3 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Food choices are complex and are affected by a combination of various factors, including biological determinants (e.g., hunger, appetite, and taste), psychological determinants (e.g., mood, stress, and guilt), physiological determinants (e.g., access, education, and time), social determinants (e.g., culture, family, and peers), and economic determinants (e.g., cost, income, and availability). Attitudes, beliefs, and knowledge about food also have an influence on food choices [5]. However, these factors could affect people differently, depending upon their context, personality, social groups, and socio-cultural position.

In the literature it has been discussed that health, mood, convenience, price, familiarity, social norms, natural and ethical concerns, and taste are prime issues considered by consumers when making food choices [6–10]. Eertmans et al. (2006) conducted a survey using the food choice questionnaire (FCQ) in different countries and found that in Italy, health and nature content, convenience and mood were the most important issues Italian consumer concerned in making their food choices [11]. Guiné et al. (2019) studied food choice determinants in Mediterranean countries and found that, in Italy, the food choices were influenced by environmental and political motivation, following by health and emotional reasons [12].

In order to determine the eating patterns of individuals in relation to their choices, particularly in the Mediterranean region, a large project entitled "Psycho-social motivations associated with food choices and eating practices" (EATMOT) was carried out. In this framework, a questionnaire was developed to define food choices, according to six types of conditioning motivations (health, emotional, economic and availability, social and culture, environment and politics, and marketing and commercial) [12]. This study was part of the project and its purpose was twofold: (1) To evaluate the food choice motivations in a sample in Italy; and (2) to segment consumer groups and provide consumer profiling, according to their food choice motivations.

#### **2. Materials and Methods**

#### *2.1. Questionnaire*

A questionnaire was developed specifically for the EATMOT project by the Center for Education, Technology and Health Studies (CI & DETS Research Center) in Portugal [13]. The questionnaire was prepared in English, then translated into Portuguese for the pre-test and validation of the questionnaire before the actual survey was carried out in 16 countries. The initial scale validity and internal reliability of the questionnaire were assessed only in Portugal (i.e., for the Portuguese version; see details in [13]). After validation, the questionnaire was modified and subsequently translated into English. The final version of the questionnaire was translated into Italian, following a back-translation methodology for validation [14]. During the translation process, the questions were slightly adjusted, in order to be more coherent with Italian culture, while the original meanings were retained. The questionnaire structure included five sections: Part 1—Socio-demographic data; Part 2—Anthropometric data and behavioral- and health-related elements; Part 3—Attitudes relating to healthy food; Part 4—Sources of information about a healthy diet; and Part 5—Food choice motivations (M1: Healthy motivations, M2: Emotional motivations, M3: Economic and availability motivations, M4: Social and cultural motivations, M5: Environmental and political motivations, and M6: Marketing and commercial motivations). The individual items and type of scale for all measures are provided in Appendix A (Table A1).

The questionnaire comprised both closed- and open-ended questions. In the perception of healthy eating and the food choice motivation sections, respondents were asked to give their opinion toward statements according to a 5-point Likert-like scale, ranging from 1 (Strongly disagree) to 5 (Strongly agree); while in the sources of information section, participants were asked to indicate the frequency at which they found information about healthy diets from different sources, on a scale from 1 (Never) to 5 (Always).

#### *2.2. Data Collection*

The questionnaire was administered both offline and online (Google forms) using a convenience sample of the Italian population through the personal connections of the authors. The online survey link was distributed through personal emails. Offline data collection was conducted in the North and Central parts of Italy. The interviews were carried out face-to-face with randomly selected consumers in different parts of the town (e.g., grocery stores and supermarkets) by experienced researchers/graduate students under the supervision of the authors of this paper. Data was collected between January and September 2017. The target respondents were adults aged over 18 years old, who voluntarily provided their consent to participate in the study.

All ethical procedures were strictly followed when designing and applying the questionnaire, and it was ensured that the data provided were kept strictly confidential (i.e., such that no individual response could ever be associated with the respondent). The study was conducted in accordance with the Declaration of Helsinki and the protocol was approved by the Ethics Committee of Polytechnic Institute of Viseu (reference nº 04/2017); furthermore, national and international protocols for research on humans were followed.

In total, 585 individuals participated in the survey. Through the validation process (i.e., elimination of incomplete questionnaires, leaving out outliers, and replacing invalid values with the mean), 531 questionnaires were considered valid and used in data analysis phase.

#### *2.3. Data Analysis*

Data were analyzed through both univariate and multivariate techniques using the IBM SPSS 26.0 software. A basic descriptive approach was used to describe Italian consumer characteristics, in terms of socio-demographics, anthropometrics, health-related behaviors, and information sources about healthy diets, including perceptions about healthy eating. Body mass index (BMI) was calculated using self-reported height (m) and weight (kg) data. The BMI results were classified according to International Classification Standards [15], as follows: underweight (BMI < 18.50 kg/m2), normal weight (18.50 ≤ BMI ≤ 24.99 kg/m2), overweight (25.00 ≤ BMI ≤ 29.99 kg/m2), and obese (BMI ≥ 30.00 kg/m2).

In the food choice motivations section, the median, mean, and standard deviation (SD) values of each item were calculated. Note that, to calculate the global scores, the inverted scores of M1.5 ("There are some foods that I consume regularly even if they may raise my cholesterol"), M1.9 ("There are some foods that I consume regularly even if they may raise my blood glycaemia"), M4.5 ("I prefer to eat alone"), M6.1 ("When I buy food I usually do not care about the marketing campaigns happening in the shop"), and M6.4 ("When I go shopping I prefer to read food labels instead of believing in advertising campaigns") were used, as they were negative questions (according to the motivations). Hence, the higher the global score, the stronger the influence of the motivations on food choices.

Consumer groups were identified using data of the food choice motivations section. Cronbach's alpha was used to test the internal validity of the 49 food motivation items (Cronbach's α = 0.755). Then, all 49 items from the six food choice motivations were used for consumer segmentation (10 healthy items, 9 emotional items, 7 economic and availability items, 9 social and cultural items, 7 environmental and political items, and 7 marketing and commercial items). First, Hierarchical cluster analysis (HCA) with Squared Euclidean distances and Ward's method was applied to the items, in order to define the optimum number of clusters. The agglomeration schedule suggested that 2 clusters were suitable for the collected data. Then, K-mean cluster analysis was applied to identify the final clusters. Finally, the resulting clusters were evaluated, according to socio-demographics, anthropometrics, health-related behaviors, information sources, and perception about healthy eating, using the Pearson Chi-square, Student t-tests, and Mann–Whitney U test for independent samples.

#### **3. Results**

#### *3.1. Sample Characteristics*

3.1.1. Socio-Demographic and Anthropometric Data and Behavioral and Health Aspects

The sample included 531 Italian participants, of whom 65% were female. Their age ranged from 18 to 75 years, with a large group of respondents aged between 35 and 44 years old (27.7%), followed by respondents aged between 25 and 34 (23.9%). The average age was 42 years (SD = 13.47). The majority of respondents were higher educated and held a university degree (50%), while 45% possessed a secondary school diploma and around 5% of respondents had completed primary school. In terms of residences, 61% of the respondents lived in urban areas, 29% lived in suburban areas, and 10% lived in rural areas. Considering their civil status, 66% of respondents were married/living together, while 26% were single, 6% were divorced or separated, and 2% were widowed. The majority of respondents were employed (56%), while 19% were housewives, 9% were retired, 6% were unemployed, 6% were students, and 4% were working students. Most respondents were responsible for buying food for their household (85%).

Based on the self-reported weight and height, the majority of respondents were of normal weight (63.8%), whereas overweight and obese individuals comprised 24.5% and 5.8% of the study sample, respectively. Two hundred responders (37.7%) described themselves as being physically active, 54.8% of the sample declared having a healthy diet, and 76% of responders were not dieting or following a particular dietary regimen. A total of 72.3% of participants declared not having chronic diseases, while only a few suffered from allergies and/or intolerances (16.6%) or experienced eating disorders (9.2%).

#### 3.1.2. Information Sources

Respondents were asked to indicate the frequency at which they found information about eating a healthy diet from different sources of information. They frequently used the internet and magazines, books and newspapers, and sometimes family or friends, television, and doctors. They found information at school or on the radio sporadically (Table 1).


**Table 1.** Sources of information for the total sample (*n* = 531).

Note: Respondents were asked to indicate the frequency at which they found information about eating a healthy diet, on the following scale: 1 = never, 2 = sporadically, 3 = sometimes, 4 = frequently, 5 = always.

#### *3.2. Perceptions about Healthy Eating*

The median and average scores of the respondent's perceptions about healthy eating are displayed in Table 2 and Figure 1. Almost all respondents strongly agreed that a healthy diet should be balanced, varied, complete, and should include fruit and vegetables. They also agreed that it is important to eat everything, although in small quantities. Disagreement was observed for inverted questions related to totally avoiding sugary and fatty products and having cravings for sweets, for some people. For questions related to the price of a healthy diet, the value of tradition for healthy patterns, a healthy diet being based on calorie count, or organically produced foods being healthier than their

conventional counterparts, the responders neither agreed nor disagreed; in fact, their scores were quite variable.



Note: Respondents were asked to indicate their opinion on the statements, based on a 5-point semantic scale: 1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree.

**Figure 1.** Box plot showing perception scores toward healthy diet (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree).

> We tested the correlations between the different variables recorded in Part 2 of the questionnaire and perceptions toward healthy eating. The following results are those that were correlated and differed among groups at a significance level of 95%. A Kruskal–Wallis test provided strong evidence of a difference (*H* = 9.866, *p* = 0.02, df = 3) among the BMI classes toward the statement "A healthy diet should be balanced, varied and complete". The Dunn–Bonferroni post-hoc method was carried out. Normal weight respondents rated the highest score for this statement, and there was evidence that it was significantly higher than those of obese (*p* = 0.05) and underweight (*p* = 0.015) respondents, which indicates that this statement could be key to staying healthy and maintaining a normal weight.

Underweight respondents rated the statement "We can eat everything, as long as it is in small quantities" significantly lower than normal weight (*p* = 0.013), overweight (*p* = 0.007), and obese (*p* = 0.002) respondents. Underweight respondents also rated the statement "We should never consume fat products" significantly lower than overweight (*p* = 0.028) and obese (*p* = 0.008) respondents. These results meant that underweight respondents were very concerned about the types of food they consumed, but they disagreed that fatty products should be avoided, while overweight and obese respondents were very concerned about consuming fatty products.

The Mann–Whitney U-test was used to assess the following relations. Respondents who stated that they frequently/always followed a healthy diet scored the statements "Fruit and vegetables are very important to healthy eating" (*z* = −2.998, *p* = 0.003) and "A healthy diet should be balanced, varied and complete" (z = −3.193, *p* = 0.001) significantly higher than those who reported that they did not follow a healthy diet. Respondents who stated that they moderately/intensively did physical activities scored the statement "A healthy diet should be balanced, varied and complete" significantly higher than those who did not (*z* = −2.597, *p* = 0.009). Respondents who had chronic diseases scored significantly higher than those who had not on the statements "We should never consume sugary products" (*z* = −2.225, *p* = 0.026) and "We should never consume fat products" (*z* = −2.956, *p* = 0.003), indicating that chronic disease affects the perception of a healthy diet. Respondents who had experienced an eating disorder scored the statement "In my opinion, it is strange that some people have cravings for sweets" significantly lower than those who had never experienced one (*z* = −2.088, *p* = 0.037), showing that, for those who had ever experienced an eating disorder, cravings for sweets were normal. Respondents who followed a voluntary food regimen rated the following statements lower than those who did not follow any food regimen: "We should never consume sugary products" (*z* = −4.931, *p* < 0.001), "We should never consume fat products" (*z* = −3.143, *p* = 0.002), and "I believe that organically produced food is healthier" (*z* = −2.741, *p* = 0.006); while they rated the statement "I believe that tradition is very important to a healthy diet" significantly higher than those who did not follow any food regimen (*z* = −3.432, *p* = 0.001).

In summary, the statistical analysis demonstrated that BMI class and being active significantly affect the perception of a healthy diet as balanced, varied, and complete. Having a chronic disease significantly affects perceptions related to avoiding fatty and sweet products. Respondents who followed a voluntary food regimen had significantly different perceptions about a healthy diet than those who did not follow any regimen, regarding fat and sugar consumption, the role of tradition, and the healthiness of organically produced foods.

#### *3.3. Food Motivations*

The items associated with food choice motivations are shown in Figure 2 and Table 3. Environmental and political motivations (mean = 3.64, SD = 0.57), as well as health (mean = 3.4, SD = 0.46), were the strongest determinants (see Figure 2); while social and cultural (mean = 3.07, SD = 0.36), emotional (mean = 2.96, SD = 0.67), and economic and availability (mean = 2.86, SD = 0.51) motivations were less considered by respondents. Marketing and Commercial motivations were considered the least important drivers of food choices (mean = 2.46, SD = 0.53).

Twenty-three out of the 49 items had a median equal to 4 (median respondents agreed with the statements; see Table 3). It is worth noting that the non-inverted scores of statements M1.5, M1.9, M6.1, and M6.4 (negative statements, according to their global motivations) are shown here as well, as they were actually rated as 4 or above (i.e., agree). Respondents agreed about the social nature of meals ("Meals are a time of fellowship and pleasure"). They preferred to read food labels, instead of believing in marketing and commercial ("When I go shopping I prefer to read food labels instead of believing in advertising campaigns"). They also cared about the quality of their diet, in order to stay healthy ("It is important for me to eat food that keeps me healthy"), and about

environmental sustainability ("When I cook I have in mind the quantities to avoid food waste", "It is important to me that the food I eat is prepared/packed in an environmental friendly way"). Overall, the respondents were very concerned about the environmental and health aspects of their food choices; nevertheless, they also considered emotional ("Food makes me feel good"), economic ("I usually choose food that has a good quality/price ratio"), and social ("I eat more than usual when I have company") aspects of food. Although the health aspect was crucial for them, they also regularly consumed some foods that may raise their cholesterol and blood glycemia ("There are some foods that I consume regularly, even if they may raise my cholesterol", "There are some foods that I consume regularly, even if they may raise my blood glycaemia").

**Figure 2.** Average motivation scores of total sample (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree).

#### *3.4. Consumer Segmentation*

The items associated with the food choice motivations and results from the cluster analysis are shown in Figures 3 and 4. Two clusters were identified: Cluster 1 "Emotional eating consumers" and Cluster 2 "Health-driven consumers". Regarding the most and least important motivations, both clusters had the same idea: "Environmental and Policy motivations" were the most important, while "Marketing and Commercial motivations" were the least considered. Details of the scores of each item for each motivation, including the statistical difference between Clusters, are given in Appendix A.


**Table 3.** The most important food choice motivations for the total sample (*n* = 531).

Note: Respondents were asked to indicate their opinion on the statements based on a 5-point semantic scale: 1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree. Median scores of all items were equal to 4 (agree).

> The first cluster accounted for 54.24% (288 persons) of the total sample and was described as "Emotional eating consumers" (Figure 3). Besides "Environmental and Policy motivations" (mean = 3.56, SD = 0.52), "Emotional motivations" (mean = 3.36, SD = 0.52) were very important for this cluster, as they scored most emotional items higher than the respondents in Cluster 2 (*t* = 19.529, *p* < 0.001). Food helped them to cope with stress, made them feel good, and served as their emotional consolation. In addition, they tended to emotionally eat, as they ate more when they felt lonely or had nothing to do, including craving sweets when they were depressed. They also consumed food to either keep them alert or relax. "Health motivations" (mean = 3.29, SD = 0.46) were the third most important for them; however, they scored most items in this category lower than respondents in Cluster 2, while they scored higher regarding consuming some foods regularly, even if they may raise their cholesterol or blood glycaemia (inverted scores). They also cared for "Social and cultural motivations" (mean = 3.15, SD = 0.37), "Economic and availability motivations" (mean = 3.09, SD = 0.44), and "Marketing and commercial motivations" (mean = 2.70, SD = 0.44) more than respondents in Cluster 2 (*t* = 5.391, *p* < 0.001; *t* = 13.092, *p* < 0.001; and *t* = 12.998, *p* < 0.001, respectively).

**Figure 3.** Average motivation scores of Cluster 1 (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree).

**Figure 4.** Average motivation scores of Cluster 2 (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree).

The second cluster was named "Health-Driven consumers", which accounted for 45.76% (243 persons) of the total sample. While "Environmental and Policy motivations" (mean = 3.73, SD = 0.61) were the most important drivers for them, "Health motivations" (mean = 3.54, SD = 0.43) were also highly considered. They agreed that they usually followed a healthy and balanced diet, it was important for them to eat foods that kept them healthy, and they tried to eat foods that did not contain additives. In addition, considering environmental motivations, they preferred to eat food that was prepared/packaged in an environmentally friendly way, and they were concerned about food waste and the reduction of food packages, differing from the respondents in Cluster 1. They cared for the environment more than the respondents in Cluster 1 (*t* = −3.500, *p* = 0.001). Besides those two motivations, "Social and cultural motivations" (mean = 2.98, SD = 0.34), "Economic and availability motivations" (mean = 2.59, SD = 0.44), "Emotional motivations" (mean = 2.49, SD = 0.50), and "Marketing and commercial motivations" (mean = 2.18, SD = 0.49) were less concerned by them.

#### *3.5. Consumer Profiling*

In order to understand the differences between the two segments, Pearson Chi-square, Student T-test, and Mann–Whitney U-test were performed on their demographic and anthropometric data and behavioral and health aspects. The results revealed that age, life environment, behavior, and health-related elements could significantly differentiate the segments (Table 4).




**Table 4.** *Cont*.

Note: *p*-values were results from Pearson Chi-square, except a, which resulted from a Student's T-test and b, which resulted from a Mann–Whitney U-Test between the two Clusters.

> The average age (44 years old) of respondents in Cluster 2 (or Health-Driven consumers) was significantly higher than that (40 years old) of Cluster 1 (or Emotional eating consumers; *t* = −3.362, *p* = 0.001). A significantly lower percentage of respondents in Cluster 2 (55%) lived in urban areas than those in Cluster 1 (66%; *χ<sup>2</sup>* = 7.936, *p* =0.019). Consistent with their scores for items in Health motivations, 69% of respondents in Cluster 2 stated that they followed a healthy diet. This was significantly higher than that in Cluster 1 (43%; *χ<sup>2</sup>* = 35.059, *p* < 0.001). The respondents in Cluster 2 (46%) also physically exercised more than respondents in Cluster 1 (31%; *χ<sup>2</sup>* = 12.256, *p* < 0.001). Respondents in Cluster 2 (30%) followed a voluntary food regimen more than those in Cluster 1 (24%; *χ<sup>2</sup>* = 8.639, *p* = 0.003). The clusters were also differentiated—although with lower significance (statistically significant at 0.1 level)—in terms of the following issues: Based on BMI categories,

a higher number of respondents in Cluster 2 (68%) had normal weight than in Cluster 1 (61%; *z* = −1.839, *p* = 0.066). However, more respondents in Cluster 2 (31%) had chronic diseases than in Cluster 1 (25%; *χ<sup>2</sup>* = 2.888, *p* = 0.089), likely due to their higher average age. Additionally, the respondents in Cluster 1 (11%) had more experiences with eating disorders than those in Cluster 2 (7%; *χ<sup>2</sup>* = 3.738, *p* = 0.053).

Regarding food motivations, respondents in Cluster 1 agreed with 19 out of 49 items (i.e., median respondents agreed with the statements; see Table 5). The non-inverted scores of statements M6.4, M4.5, M1.5, and M1.9 (negative statements, according to their global motivations) are shown here, as they were actually rated as agree. Respondents agreed that food had emotional value for them ("Food makes me feel good"). They also cared about the health ("It is important for me to eat food that keeps me healthy") and environmental ("When I cook I have in mind the quantities to avoid food waste") aspects of food. Although they were concerned about economics ("I usually choose food that has a good quality/price ratio"), they preferred to read food labels, instead of believing in advertisements ("When I go shopping I prefer to read food labels instead of believing in advertising campaigns"). Some preferred to eat alone ("I prefer to eat alone"). In general, the respondents in Cluster 1 concerned many aspects (motivations) of food, compared to respondents in Cluster 2. Emotional motivations seemed to be very important to them ("Food makes me feel good", "I eat more when I have nothing to do", "I have more cravings for sweets when I am depressed", and "Food helps me cope with stress").

Respondents in Cluster 2 agreed with 15 out of 49 items (i.e., median respondents agreed with the statements; see Table 6). The non-inverted scores of statements M6.1 and M6.4 (negative statements, according to their global motivations) are shown here as well. Respondents cared the most about health ("It is important for me to eat food that keeps me healthy") and environmental ("When I cook I have in mind the quantities to avoid food waste" and "It is important to me that the food I eat is prepared/packed in an environmental friendly way") aspects of food. They also agreed with social and cultural aspects ("I choose the foods I eat, because it fits the season"). Similar to the respondents in Cluster 1, they preferred to read labels, rather than believing in commercial advertisements ("When I go shopping I prefer to read food labels instead of believing in advertising campaigns"). Generally, the respondents in Cluster 2 were more concerned about health aspects than respondents in Cluster 1 ("It is important for me to eat food that keeps me healthy", "Usually I follow a healthy and balanced diet", "It is important for me that my daily diet contains a lot of vitamins and minerals", and "I try to eat foods that do not contain additives").

Regarding perceptions about healthy eating, there were also differences on perception about healthy diet between the two clusters (Table 7). Cluster 1 agreed significantly more than cluster 2 with the importance of eating everything (although in small quantities), with the role of tradition in a healthy diet, and that a healthy diet is not cheap. Differences between the clusters about totally avoiding sugary and fatty products were significant at the 0.10 level.

The frequency of finding information about eating a healthy diet from different sources of information is shown in Figure 5. Respondents in Cluster 2 used "specialized" sources, such as schools, health centers, hospitals, and family doctors (GP) more than respondents in Cluster 1. On the contrary, respondents in Cluster 1 used mass media—both traditional (radio and television) and internet—as well as books, magazines, and word-ofmouth between family and friends more than respondents in Cluster 2. All these sources, moreover, are cheaper than consulting experts (i.e., doctors).


**Table 5.** The most important food choice motivations for Cluster 1 (*n* = 288).

Note: Respondents were asked to indicate their opinion on the statements, based on a 5-point semantic scale (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree). Median scores of all items were equal to 4 (agree).


#### **Table 6.** The most important food choice motivations for Cluster 2 (*n* = 243)


#### **Table 6.** *Cont*.

Note: Respondents were asked to indicate their opinion on the statements, based on a 5-point semantic scale (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree). Median scores of all items are equal to 4 (agree).


**Table 7.** Differences between clusters on perceptions about healthy eating.

Note: Respondents were asked to indicate their opinion on the statements based on a 5-semantic scale (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree). The *p*-values are the result of Mann–Whitney U-Test between two clusters.

> In order to understand the differences between the information sources used by the two segments, a Mann–Whitney U-test was performed (Table 8). The results revealed that trust in TV and Radio significantly differentiated the segments. The respondents in Cluster 1 had more experience with using Radio and Television for information about healthy eating than those of Cluster 2.

**Table 8.** Difference in information sources about healthy diets between clusters.


Note: Respondents were asked to indicate the frequency at which they found information about eating a healthy diet, on the following scale: 1 = never, 2 = sporadically, 3 = sometimes, 4 = frequently, 5 = always. The *p*-values are the results of Mann–Whitney U-Test between two clusters.

**Figure 5.** Average scores of information sources (1 = never, 2 = sporadically, 3 = sometimes, 4 = frequently, 5 = always).

#### **4. Discussion**

In this study, we investigated the motivations behind food choices in Italy and segmented the surveyed consumers to provide recommendations on effective tools to encourage healthy food choices. The main results indicated that some factors influenced consumer food choices more than others; for instance, in line with the previous literature [11,14,16–19], "Environmental and Political" and "Health" motivations were the most important determinants of food choices for Italian consumers, while "Marketing and Commercial" motivations were of the least concern [20].

When we looked at individual items, the highest rated items were as follows: Respondents mostly agreed that meals are linked to companionship and pleasure. This might be explained by the Italian tradition that meal time is a time to spend with family and friends [21]. They also stated that they preferred to read food labels when they shop for food, instead of believing in marketing campaigns. It was also highly important for them to eat food that keeps them healthy. Food waste was also highly considered by respondents when they prepare food, which was consistent with the results of Bravia et al. (2020): That Italian consumers tend to be proactive in planning their food purchases and checking 'use by' and 'best before' dates of food products, such that they can reduce their household food waste [22]. It has also been mentioned, in a report of the European Union, that trying to reduce waste was the number one action Italians carried out to address the issue of climate change [19].

Two consumer clusters were identified, according to the six analyzed food motivations: (1) "Emotional eating consumers", composed of respondents that were driven by their emotions; and (2) "Health-driven consumers" composed of those that based their choices on the health aspect of food. Although both clusters were primarily concerned with "Environmental and Political" motivations, when it came to food choice, their second-most important motivations differed; namely, "Emotional" and "Health", respectively.

The food choices of Emotional eating consumers were affected by psychological factors (e.g., mood). However, they also paid attention to health, food waste, food labels, and economic and availability factors of the product. Basically, they considered more aspects than Health-driven consumers. However, their choices could be highly affected by their mood or emotional state, as they mentioned that food made them feel good, that they ate more when they had nothing to do, that they had more cravings for sweets when they were depressed, and that food helped them to cope with stress. Their average age was lower than that of Health-driven consumers. Our results were in line with those of Cardoso et al. (2020), who stated that young adults had eating behaviors which were more conditioned by emotional motivations (e.g., to fight loneliness or boredom), compared to senior adults and elderly people [23]. Moreover, young adults have also been reported to link eating food with emotional consolation; for example, to help deal with stress and negative moods [24]. In addition, Emotional eating consumers had more experiences with eating disorders than Health-driven consumers, as they tended to exhibit emotional eating behaviors. Emotional eating consumers also agreed that there was some food that they consumed regularly, even if it may raise their cholesterol/blood glycaemia.

The food choices of Health-driven consumers were mainly driven by health-related aspects, seasonal availability, and label information. The relevance of sustainable consumption movements was also highlighted in this cluster. Besides avoiding creating food waste, similarly to Emotional eating consumers, they also believed that food should be prepared/packaged in an environmentally friendly manner, using minimal packaging. For them, it was also important that the foods they consume came from countries where human rights are not violated. When we looked at their profiles, the average age of Health-driven consumers was significantly higher than Emotional eating consumers. As a consequence, they had more chronic diseases than Emotional eating consumers. Health-driven consumers also stated that they followed a healthy diet and/or voluntary food regimen and exercised more than Emotional eating consumers. This is consistent with the fact that elderly individuals tend to eat more fruit and vegetables and are usually more adherent to healthy diets than young adults [25]. The fact that they have more chronic diseases may drive them to adhere to dietary recommendations and to be active. As a consequence, they tended to have a normal weight more than Emotional eating consumers.

Regarding perceptions related to healthy eating, most respondents perceived that a healthy diet should be balanced, varied, and complete and should include fruit and vegetables, which are aspects in accordance with suggestions for healthy eating by nutritionists [26–29]. This demonstrated that they were mostly aware about what healthy foods are. Nevertheless, only half of them stated that they followed a healthy diet. Hence, there existed a gap between declarative knowledge and behaviors. Several studies have confirmed that healthy eating knowledge is a significant predictor of both future knowledge and behavior [30–33]; however, knowledge alone is not sufficient to change the food behaviors of consumers, as such behaviors can also be influenced by personal, intra-individual, and environmental factors, including motivations [34–36]. The results of our cluster analysis underline that one such factor could be "Emotional" motivations; at least, for Emotional eating consumers. In addition, unconscious motivations and the link between nutritional knowledge, emotions, and food choice should be further investigated.

Most respondents also agreed that they could eat everything, as long as it is in a small quantity. In addition, they disagreed about the "total avoidance of sugary and fatty products". This may be because they believed that small quantities and variety are key to a healthy diet, as suggested by the Italian Food Dietary Guideline [37]. Moreover, they may have taken into account the emotional effect on food choice, as they generally agreed that having a craving for sweets is not unusual.

Respondents who had a normal body weight, declared to have healthy diet, and were active believed that a healthy diet should be balanced, varied, and complete. Overweight and obese respondents were more concerned about the consumption of fatty products than other groups (i.e., agreeing more that we should avoid consuming fatty products). Respondents who had a chronic disease perceived that they should avoid fatty and sweet products, as suggested in the various guidelines for the treatment of heart disease and for preventing dietary diseases in the general population (see, e.g., the Diet and Lifestyle Recommendations by the American Heart Association) [38].

In addition, following a voluntary food regimen led respondents to perceive healthy eating differently than those who did not follow any voluntary food regimen, with respect to fat and sugar consumption, the role of tradition, and the healthiness of organically produced foods. This might be because individuals who follow a particular diet are usually more attentive to their diet and are more informed about their diet, such that they are more aware of what should be limited (e.g., fat and sugar). They also considered that organic products are produced with more respect for the environment [39]. Especially in Italy, organic products are considered not only environmentally friendly but healthy; although, from a nutritional point of view, organic products are not considered to differ much from non-organic ones [39–41].

Concerning the sources of information used, our sample was well-informed and took information from various sources. The internet was the main source of information about healthy eating, followed by magazines, books, and newspapers. These answers were consistent with the emerging informative use of the Internet (i.e., reading online newspapers, documenting health, collecting information on products), although the Italian people are still in the last position in Europe, in terms of informative usage of the Internet, at present [42]. The internet also consists of the use of social networks. The use of social networks has been shown to change the way that consumers search for information and select products; they are becoming prominent sources of information, including for food choices (see [43,44], for example). From this perspective, social media can influence information strategies in two ways: Reducing the cost of releasing information, compared to that of traditional mass media (i.e., television or radio), and making specific consumer groups more easily targeted.

Offline and online word-of-mouth and social media, however, can be dangerous in transmitting misinformation; especially in the food sector [45,46]. Indeed, the recent study of Castellini et al. (2020) found that around half of Italians (48%) admitted that they had believed in a news story about the food sector that turned out to be false at least a few times in the last year, while a third of those (37%) had shared it on social media, thus contributing to the unstoppable spread of "food fake news" [46]. An even more interesting fact is that this phenomenon occurs in all social and educational classes. In particular, individuals who believe in such misinformation are psychologically different than consumers who are less persuaded by this kind of news. They are more driven by other motivations, related to familiarity with the product and the mood of the moment, rather than by the evaluation of healthiness of the food. They are more favorably oriented to experimentation with new products and are more predisposed to social influence, being less self-confident [46]. Hence, the Emotional eating consumers could be at risk of being more susceptible to such misinformation. This means that the role of institutions in educating the public is important, and that they should exploit different media forms, in order to aid citizens to be able to distinguish between reliable and non- reliable sources of information and enable them to make well-informed food choices.

For policy-makers, food labels could be promoted as a tool to assist consumers to make a conscious food choice. Information provision to promote sustainable and healthy diets could be carried out through educational campaigns (e.g., relating to the inclusion of sustainability in dietary guidelines) or improved sustainability labelling on packaging. Social media could be used to change social norms and food culture towards healthy diets and waste reduction, as recent studies have demonstrated that social media information also affects environmental awareness and consumer information and choices relating to sustainable food [44]. Further research is necessary to examine the role of the Internet in food information more closely, as well as the sources which are judged as most reliable by respondents. It will also be interesting, in future research, to deepen the multifaceted relationship between traditional and social media information and healthy or sustainable consumption, in terms of food choices for specific consumer segments.

For food marketers, there was a clear need, for both clusters, for food products that are healthy and sustainable. Environmentally friendly packaging and human rights in the producing countries were also emphasized by Health-driven consumers. Moreover, information regarding the environment and sustainability on these issues (e.g., product labelling) should be provided, in order to increase the purchase of such products [47].

The limitations of this research were as follows: There was a higher proportion of female participants and the sample had higher education, on average, than the Italian census. BMI values were calculated from self-reported height and weight and, therefore, the results should be interpreted with care. We also did not ask respondents about their dietary habits or their consumption of some food products, such that we could not compare their statements (e.g., following healthy diet) and (reported) consumption behavior. Future research should cover this aspect.

#### **5. Conclusions**

In conclusion, the respondents had a good awareness of what a healthy diet consists of. They mostly found information online or in newspapers and books, as well as through talking with friends and families. The strongest determinants for their food choice were Environmental and Health factors. The less influential reasons were those related to Marketing and Commercial. The clustering analysis resulted in two consumer segments: Emotional eating consumers and Health-driven consumers. Both segments considered Environmental and Political motivations as the most important issues. Nevertheless, their second-most important motivations divided them, as the Emotional eating consumers were more influenced by their emotions, while Health-driven consumers were more concerned with the health aspects of food. Emotional eating consumers were younger, while Healthdriven consumers had more normal weight and stated that they followed a healthy diet and/or voluntary food regimen and exercised more than Emotional eating consumers. Food labels were used by respondents as an important tool when making food choices. Food waste and food packaging were issues also concerned by most respondents.

**Author Contributions:** Conceptualization, R.P.F.G., C.M., N.P. and E.V.; methodology, C.M. and R.W.; data collection: C.M., N.P., E.V. and E.C.; formal analysis, R.W.; writing—original draft preparation, R.W., C.M. and N.P.; writing—review and editing, R.P.F.G., G.S., N.P., E.C. and E.V. All authors have read and agreed to the published version of the manuscript.

**Funding:** The APC was funded by Parma University.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of Polytechnic Institute of Viseu (reference nº 04/2017).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The datasets generated for this study are available on request to the corresponding author.

**Acknowledgments:** This work was prepared in the ambit of the multinational project EATMOT from CI&DETS Research Centre [IPV—Viseu, Portugal] with reference PROJ/CI&DETS/CGD/0012. This work is supported by Portuguese National Funds through the FCT—Foundation for Science and Technology, I.P., within the scope of the project Refª UIDB/00681/2020. Furthermore, we would like to thank the CERNAS Research Centre and the Polytechnic Institute of Viseu for their support.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Appendix A**

Food choice motivation of total sample, Cluster 1 and Cluster 2.


**Table A1.** Food choice motivation of total sample, Cluster 1 and Cluster 2.


**Table A1.** *Cont*.


**Table A1.** *Cont*.

Note: Respondents were asked to indicate their opinion on the statements based on a 5-point semantic scale (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree). <sup>a</sup> Inverted scale as they were negative questions (according to the motivations), *p*-values are results from Mann–Whitney U-Test between the two clusters.

#### **References**

