Abstract
Background: Cultural capital can interestingly mediate what a person prefers to eat and, thus, may be effective in dietary choice. The objectives of this study were to compare the distribution of cultural capital components within the different educational levels and also compare food consumption patterns across the different educational levels; and examine cultural capital’s role in the educational inequalities in food consumption among healthy women. 
Methods: The data was obtained in the form of a cross-sectional design with face-to-face interviews with 527 women who had visited shopping centers and parks in Tabriz from September to November 2021, using convenient sampling. Food recall as well as cultural capital questionnaire were completed for the participants, and anthropometric indices were measured. The level of education of the participants was considered as an indicator of socio-economic status. 
Results: Family institutionalized cultural capital, objectivized cultural capital, and most of the incorporated cultural capital’s indicators were significantly differed by educational inequalities. Daily frequency scores of unhealthy food consumption in the participants with primary education was higher than other educational groups (p<0.001). The high-educated individuals’ intakes healthier foods were compared to unhealthy foods, and participants with low total cultural capital score were in adherence of a low overall healthy food consumption (PR: 1.49, 95%CI: 1.23-1.68). 
Conclusion: Education levels are significantly associated with cultural capital, and participants with high levels of cultural capital choose healthier foods more often than participants with low cultural capital. 
Keywords: Cross-sectional studies, Diet, Economic status, Educa-tional status, Female, Humans, Surveys and questionnaires

Introduction
Dietary choices are a main indicator of population health, morbidity, and mortality (1). Healthy diets which are high in fruit and vegetables, whole grains, nuts, and low-fat dairy products may be protective in incidence of Non-Communicable Diseases (NCDs) (2). Conversely, consumption of unhealthy diets such as foods with high amount of saturated fatty acids, sugars, and salt which are poor in micronutrients has contributed to most of the chronic disorders such as diabetes, coronary heart disease, hypertension, and some cancers (3,4). 
In addition, over-consumption of energy-dense diets are linked to the development of overweight and obesity, which in turn are associated with much co-morbidity (5-7). It is estimated that worldwide prevalence of obesity will grow by 9.7 billion, in 2050 (8). The sharp rise in incidence of obesity contributes to westernization of the diet and also household incomes that can influence the dietary choices (9). Further, socioeconomic position, living conditions, and cultural resources partly mediate the food choice behaviors (10), and socio-economic inequalities lead to discrepancies in health status (11). Many of the recent national dietary surveys demonstrated that healthier dietary intake is generally associated with higher social and cultural status (12-14). In fact, unhealthy behaviors are not the result of completely voluntary choices of individuals, but are more influenced by the structures in daily life such as material resources and facilities. Low socio-economic groups usually have unhealthier behaviors such as unhealthy food choices (15,16). Educational level is a strong and commonly-used indicator of socio-economic status that can predict consumption of a healthy diet, as nutritional knowledge may be more closely associated with dietary intake than traditional socio-economic characteristics (17-19). Education influences eating habits by facilitating the reading and understanding of nutritional information and adherence to nutritional recommendations (20,21). The new angle of dietary choice studies are focused on cultural resources or “capitals” that may contribute to health inequalities (22). Cultural resources can effectively influence the knowledge, norms, habits, and also preferences of population through education and socialization (23,24). According to the Bourdieu’s study, Cultural Capital (CC) is one of the distinguished forms of capital that examined in three categories including incorporated (e.g. values, skills, knowledge), objectivized (e.g. books, tools), and institutionalized (e.g. educational degrees, professional titles) (25,26). Accumulation of cultural capital in people through these three sources causes differences in those who have cultural capital and those who lack it. Bourdieu has reported that taste and food choices are influenced by cultural relations, as well as social and economic status and income. Many factors such as training, education, and attending cultural venues take part in changing lifestyle-related behaviors including food choices. Therefore, cultural capital can interestingly mediate what a person prefers to eat and, thus, may be effective in dietary choice. Some researchers consider the role of education further and consider it as “new cultural capital”. They believe that in today’s era and despite the daily changes in technology and information, the “educational capital” of individuals should be mentioned as the cultural capital of globalization. Therefore, the improvement of educational capital in professional specializations, technology and new skills can be considered as the reason for the competitive advantage of people in innovation and access to information networks, which in a way will improve the cultural capital of a person (27-29).
Relatively little is studied regarding the mechanisms underlying socio-economic inequality in eating habits or how to compensate inequalities between youth (30,31). The objectives of this study were to (1) compare the distribution of cultural capital within the different educational levels considering three dimensions of cultural capital including incorporated, objectivized, and institutionalized cultural capital (2) compare people food consumption patterns (healthy or unhealthy) across the different educational levels; and (3) examine whether cultural capital contributes to the educational inequalities in food consumption.

Materials and Methods 
Study participants
Data were obtained in a cross-sectional design by face-to-face interviews with women who were visiting shopping malls including Laleh Park, Atlas, Setareh Baran, Kourosh, and Hyper Family, as well as Elgoli and Valiasr parks in Tabriz city, Iran from September to November 2021. The inclusion criteria were women who were responsible for preparing and cooking foods at home with the age of 18 years or more. People with disabilities and who did not want to complete the questionnaires and refer for the measurement of anthropometric indicators were excluded from the study. The statistical population of this study includes all women over 18 years of age in Tabriz which is based on the latest public population and housing census. Using Cochran’s formula, the minimum sample size with a variance of 1 and an error of 0.1 is equal to 384 individuals, which is the sample floor for unlimited communities in Cochran’s formula. Considering the convenience sampling method, totally 572 subjects were participated in the study. In Tabriz, there are several famous shopping centers, parks and entertainment centers that are scattered in different parts of the city, and these centers were chosen as sample collection sites. People who had free time to visit these centers and were willing to cooperate participated in the study. The built-in chairs and tables were used for resting in the centers, as well as the tables in the food courts to make the participants to answer the questionnaires. The participation of people in the study was completely free and it was explained to the participants that their information will remain confidential for conducting this research. Also, all the participants signed and approved the informed consent form to participate in the study.

Anthropometric measurements
In order to accurately measure the anthropometric indicators, certain sections were designated as examination rooms in the centers. All the subjects were requested to visit the examination rooms after completing the questionnaires. The anthropometric indicators were measured with the help of a trained nutritionist. Anatomical measurements including height, weight, waist circumference and hip circumference were performed for all the participants. Standing height was measured by Seca height meter with an accuracy of 0.5 cm. The weight of the participants was estimated with the help of a falcon scale with minimal clothing with a measurement accuracy of 100 g. The waist circumference was measured from the middle point of the lower edge of the rib and the sternum while standing and breathing normally and with an accuracy of 0.5 cm using a tape measure. Also, hip circumference was measured using a tape measure with an accuracy of 0.5 cm and at the point that had the largest circumference. Body mass index was calculated by dividing weight (kg) by the square of height in centimeters. 

The cultural capital questionnaire deve-lopment process
The questionnaire that was used in the study of Kamphuis et al was selected for the development of cultural capital questionnaire (15). Respectively, the determinants of cultural capital that were relevant to food choice were extracted from previous studies and summarized in the form of a questionnaire. The draft of the questionnaire was then translated into Persian. The final questionnaire composed of 92 questions and was categorized into 4 sections including: 
1) characteristics of the respondents, 2) family institutionalized cultural capital that was measured by highest educational credentials of the responder’s father, mother, and partner, 3) objectivized cultural capital that was specified by availability of cooking equipment, and 4) incorporated cultural capital that was assessed by asking the participants’ participation, cooking skills, grocery shopping skills, food information skills, nutrition knowledge, and general human values. The mean score of the mentioned three dimensions of cultural capital was used as an indicator of total cultural capital, which was divided into tertiles (low, medium, and high).

Content validity
The translated questionnaire was submitted to a panel of 9 experts to review its content validity. According to the topic of the questionnaire, the expert panel consisted of public health nutritionists, dietitians, and health education specialists. The anonymous responses were collected in May and Jun 2021. The content validity was quantified using Content Validity Index (CVI) and Content Validity Ratio (CVR) (32). All the participated experts were rated every question’s accuracy, clarity, relevancy, and necessity using a 4-point Likert scale (1: very irrelevant, 2: irrelevant, 3: relevant, 4: very relevant). Some of the questions were rephrased according to the experts’ comments. Finally, the CVI and CVR scores of the questionnaire were 0.98 and 0.94, respectively, which was considered acceptable. 

Face validity
Face validity of the questionnaire was assessed by two separate phases. At first, a convenience sample of 14 adult women who were responsible in buying food items and preparing the food at home was recruited in July 2021 at Tabriz University of Medical Sciences. People who had studied in one of the fields related to nutrition or one of their close members were nutritionists, were excluded. A paper version of the finalized questionnaire was completed by the members. In the second phase, an individual structured discussion with a duration of 15 min was conducted to provide additional feedback. The strategy of using paraphrasing was used to evaluate the face validity of the questionnaire (33). As, the researcher reads the questions one by one and then asks the responder to restate the items, using their own words in response to every questions. The researchers made minor modifications to flow and clarity.

Reliability 
Reliability of the questionnaire was assessed by a pilot study consisting of a convenience sample of 34 adults. This pilot study was conducted in two stages. The printed form of the questionnaire was filled by the participants and then was re-filled by the same responders two weeks later. The responses were analyzed using SPSS Statistics software (version 20.0) to calculate Cronbach’s αlpha coefficient and Intra-Class Correlation coefficient (ICC) (34). Finally, Cronbach αlpha and ICC for the overall questionnaire were 0.90 and 0.92, respectively.

Socioeconomic status
The educational level of the participants was con-sidered as the indicator of socioeconomic status (35, 36). The educational level was measured considering three categories including: 1) primary education, 2) lower and higher intermediate general education, and 3) professional education and university.

Table 1. The characteristics of participants based on the education levels

WC: Waist Circumference, WHR: Waist to Hip Ratio, BMI: Body Mass Index. p-values obtained by One-Way ANOVA and Krauskal-Wallis Tests. p˂0.05 was considered as significant.

Table 2. Dimensions of cultural capital among the study participants according to the educational levels

* Mean and standard deviation (SD); p-values obtained by One-Way ANOVA and Kruskal-Wallis. p-value ˂0.05 was considered as significant.

Evaluation of the food intake 
The food intake of the people participating in the project was investigated by means of a 24-hr food recall. Then food intakes were analyzed by IV Nutritionist software which was modified for Iranian foods. (First Databank, San Bruno, CA, USA). To measure healthy eating of the participants, the total frequency of daily consumption of healthy foods (including fruits, vegetables, low-fat dairy products, chicken, fish, soy, and whole grains) was used. Similarly, the total frequency of daily consumption of unhealthy foods (including red meat, croissants, chocolate, white bread, carbonated beverages, and sugar) was used to obtain the unhealthy eating (37). Each of the food items were scored by their weekly consumption, as never: 0; less than once a week: 0.10; 1-2 days per week: 0.20; 3-4 days per week: 0.50; 5-6 days per week: 0.80; and every day: 1. The final scores of daily frequencies of healthy and unhealthy foods were obtained by summing the above-mentioned scores (38,39).

Statistical analysis
The SPSS version 20.0 was used to analyze the study data. Kolmogorov-Smirnov test was used to check the normality of the data. Mentioned article ways were used for scoring the cultural capital questionnaire (15). The quantitative and normal data was presented as the mean±standard deviation (SD), and in the case of non-normality and qualitative data were shown as number (percent). One-Way ANOVA and Krauskal-Wallis tests were utilized to investigate the differences of study variables and also indicators of cultural capital between subjects with various educational levels. The probable differences in daily frequency scores of the food consumption by every educational level group were obtained using Independent Sample T-test. Finally, Poisson regression models with robust variance, Prevalence Ratio (PR)’s with 95% Confidence Intervals (CI) were calculated for each of the outcomes by educational level and total cultural capital, adjusted for age, sex, Waist Circumference (WC), and marital status. 

Results 
Totally, 615 individuals were selected to participate in the study. Among them, 572 subjects had completed all parts of the study, including filling the questionnaires and anthropometric measurements. As shown in table 1, the mean±SD age of the participants were 33.80±12.43 years, which significantly differed between subjects with different educational levels, as participants with primary education and professional education levels had higher and lower age, respectively. Although a significant difference was observed in the responders’ WC between the study groups, the differences of Body Mass Index (BMI) were remained non-significant. The low-educated participants had higher WC compared to others (p<0.001). The numbers of family members decreased by advancing the level of education (p<0.001). Family institutionalized cultural capital, objectivized cultural capital, and most of the incorporated cultural capital’s indicators significantly differed by educational inequalities (Table 2). Further, the daily frequency scores of unhealthy food consumption in subjects with primary education was higher than other educational groups (p<0.001). The high-educated participants consumed healthier foods compared to unhealthy foods (Table 3). After adjusting for confounders, the participants with lower educational levels were more likely to report a low overall healthy food consumption (PR: 1.68, 95%CI: 1.47-1.88) and also high overall unhealthy food consumption (PR: 1.38, 95%CI: 1.09-1.26) (Table 4). Further, the participants with low total cultural capital followed a low overall healthy food consumption (PR: 1.49, 95%CI: 1.23-1.68). 

Table 3. Scores of daily frequency of food consumption based on the education levels

 * p<0.001, ** p<0.001, SD: Standard Deviation, p-values obtained by independent sample T-test and One-Way ANOVA. p-value ˂0.05 was considered as significant.

Table 4. Separate poisson regression models for educational level and total cultural capital in their association with food consumption (n=572)

* p<0.001, The models adjusted for the confounders including age, sex, waist circumference (WC), and marital status. p-value ˂0.05 was considered as significant.

Discussion
The results of this study revealed that low-educated participants had higher WC compared to others. Different components of cultural capital (institutionalized, objectified and most of the included cultural capital indicators) had significant differences with educational inequalities. The daily frequency scores of unhealthy food consumption have a direct relationship with education. The high-educated individuals consumed healthier foods compared to unhealthy foods. Lower educational level was more likely to be associated with a low overall healthy food consumption and also high overall unhealthy food consumption. Moreover, the participants with low total cultural capital followed a low overall healthy food consumption. The same results were reported in another study conducted in Netherlands (40). The other study among Italian adults indicated that cultural capital was more effective than social class in predicting adapting to healthy recommendations such as healthy food intake, especially among males. In this study, the relationship between cultural capital and alcohol consumption behaviors was not detected, which was also stated in a systematic-review study (41,42). Further, Perrin et al reported that consumption of more vegetables and fruits, but not high-quality fat was associated with educational level (43). It can be suggested that education, as a main part of human capital, contributes to increasing knowledge and economic status of people, which resulted in choosing a healthier diet.
Unhealthy behaviors are considered not to be the consequences of exactly voluntary choices, rather impresses by structures in the daily context. Material deprivation is a daily context that is associated with a lower and poorer household income and housing conditions. Rather financial difficulties have been demonstrated to partly mediate the relation between socioeconomic situation and health actions and manners (10,44,45). Cultural capital is certainly related to self-assessed health. People with a high cultural capital evaluate their health better than individuals with a low cultural capital (46). 
The results of the systematic review study conducted by Kamphuis et al on 113 studies showed that there is an admissible overall internal consistency and a positive relationship between cultural capital and healthy food choices, and between socio-economic status and cultural capital (15). Bourdieu has analyzed French survey data and identified different patterns of eating, leisure activities, and views of the body embodied by different social classes with varying lifestyles (47). The reduction of economic resources leads people to choose cheaper and naturally less healthy food, which in turn increases the possibility of excessive energy consumption and abdominal obesity (48). Knowledge, skills, norms, values, equipment and health promotion facilities are the key elements of cultural resources linked to health. Their promotion is important in the production and reproduction of a healthy lifestyle (49,50). Embedded cultural capital captures high-potential resources such as skills, knowledge, cognitive abilities, competencies, and aptitudes that are important for becoming health capital (46,51). The results of studies demonstrated that the more social, cultural and material resources people have, the better their health and attention to health indicators should be. The opposite of this case is also true, that is, less cultural, material and social resources are associated with less health (52-54). Education, attitudes, nutritional knowledge and cooking skills of parents, especially mothers, affect food intake and the pattern of family meals and most importantly, children (55). The study of Cembranel et al revealed that the level of education and income affects the amount of micronutrient intake, daily food consumption, low processed foods and fruits, vegetables, whole grains, milk and its derivatives and has a direct relationship with healthy food consumption (56). Regarding the transition from high school to college or university, energy consumption and unhealthy food choices decreased and healthy food consumption increased (57,58), but the transition from high school to college cultivates changes in behavior and environment that may cause weight gain (59-61). 
Previous studies in low- and middle-income countries have shown that people with less education are more active than groups with more education, and it can be a reason for their lower body mass index and waist circumference. Dietary findings indicate a positive relationship between socio-economic status and consumption of fibers, fruits, vegetables and fish. It has also been shown that groups with low socioeconomic status in high-income environments tend to use vast levels of salt and processed foods (62). High-income people spend more money on healthier foods and have more access to health care facilities to control weight gain and obesity (63-65). Previous analysis demonstrated that people with low education have less aspirations and think that they are trapped in the condition they are in. Also, low-educated people, compared to people with higher education, use their position and general material facilities less to make decisions about their life path (66). It is expected that embedded cultural capital can play an important role in the future as individuals as consumers of health not only in the contexts of their daily lives but also in clinical interactions are forced to participate more actively in determining health (46). Further, in a study among Portuguese adults, it was reported that those having more than 12 years of education compared to those with lower than 4 years adhered to a healthier diet including higher frequency of milk, vegetable soup, vegetables, fruit, and fish consumption (67). Among adolescents, along with cultural capital, other resources including social and economic capitals affected the healthy food intake (68). It is suggested to focus on the contribution of other forms of capitals in future studies to find more intricate capitals.
The food-choice specific cultural capital variables are more likely to have causal associations with food choices that is the advantage of this method. One of the strong strengths of this study is that it was carried out for the first time in Iran, which will be useful for knowing and helping the food choices of different socioeconomic levels of the population of this country. This study had some limitations that should be noted. First, due to the financial constraints, this study was conducted cross-sectionally and by random sampling method, thus the results may not be applicable to other populations. Second, although the present study was conducted in a number of parks and shopping centers selected from different parts of the city, it cannot be representative of the entire population. Further, the cultural capital questionnaire has a main limitation that is mainly specific to food. Another limitation is the self-reporting nature, which may accompany with many biases.

Conclusion
The results of this study showed that people with higher education level have significantly more cultural capital, and participants with high levels of cultural capital choose healthier foods more often than participants with low cultural capital. Cultural capital can be a new and powerful way to explain inequality in food choices, which may lead to the formulation of policies to develop interventions to promote and improve healthy food choices among low socio-economic groups. There is need for future studies to focus on the other forms of capitals to find a suitable health model.

Ethical considerations 
The study was approved by the research undersecretary of Tabriz University of Medical Sciences (Identifier: IR.TBZMED.REC.1403.668). 

Acknowledgement
The current work has been supported by a grant from
Student Research Comittee, Tabriz University of Medical Sciences (Grant number: 74571).

Conflict of Interest
The authors declare no conflict of interest.