Abstract Optimal human health largely relies on adequate nutrition, which is best achieved by consuming dietary diversity (DD) that provides all essential macro and micronutrients. This study aimed to investigate the DD and the influencing determinants among Myanmar university students in Chiang Mai, Thailand. A cross-sectional study was conducted in three universities in Chiang Mai province using a self-administered online survey from 3^rd October to 5^th November 2024. The determinants on dietary diversity score (DDS) were evaluated using multivariable analysis with multiple linear regression. A total of 170 students, 88 (51.76%) males, and 82 (48.24%) females participated in the study. The median (IQR) DDS was 6 ± 3, one score was the minimum and nine was the highest DDS. The social determinants associated with DDS included gender, religion,  self-efficacy, skipping meals due to not readiness of food, stress, eating meals with family members and others, receiving dietary suggestions from friends, limitations to diverse foods access, religious restrictions and food advertisements on billboards. Multisectoral behavior-change measures targeting different tiers of the socio-ecological model, ranging from individual to community influences, must be carried out to improve the proper dietary behaviors and diversity of the university students.

Keywords: Dietary diversity, Social determinants, Myanmar university students, Chiang Mai province, Thailand

Citation:  Kyi, P.P.N., Siewchaisakul, P., and Wungrath, J. 2026. Dietary diversity and social determinants among Myanmar university students, Chiang Mai province. Natural and Life Sciences Communications. 25(2): e2026042.

Graphical Abstract:

INTRODUCTION

Fulfilling the nutrient need would be highly possible only if numerous kinds of foods are consumed daily since no food has all the necessary nutrients (Martin-Prevel et al., 2017; Merga et al., 2022). Encouraging biodiversity and sustainability, nutritional adequacy, minimizing adverse food-related consequences, and reducing the prevalence of chronic diseases are the dimensions that people benefit from food diversity (Mukherjee et al., 2018). On the other hand, a non-diversified diet is likely to decrease physical abilities, resistance to infection, and progress in a cognitive, reproductive, and social context, thereby seriously affecting the health, growth, and well-being of individuals (Mukherjee et al., 2018). So, dietary diversity (DD) is a basic need to be achieved daily for all people throughout the lifespan. However, in low- and middle-income countries, the proportion of people who don’t meet the acceptable DD is substantially high. According to “The State of Food Security and Nutrition in Myanmar 2022-23”, 33.5% of reproductive-age women and 34.5% of children (6-23 months) in Myanmar did not get the minimum DD in 2023 (SSP, 2024). As 30.9% of adults (over 8 years) also had inadequate DD, and it was a 10.3% increase compared to data from 2022 (SSP, 2024), the prevalence of inadequate dietary diversity among the adults was significantly high.

University life is a period of high vulnerability in regarding the adverse health and dietary behaviors (Hafiz et al., 2023). A healthy diet plays a remarkable position in this critical time since that considerable population can be focused for several illness programs (Kabir et al., 2018) and their nutritional status is a true reflection of the overall health of the community (Nithya and Bhavani, 2017). Also, dietary practices during this time can shape both present and long-term eating preferences and health outcomes (Zhou et al., 2022). However, dietary intake and diversity of the university students are influenced by many factors, which include getting autonomy for food choices, affordability of food, exposure to new environment and social networks, cooking skills and facilities, nutritional knowledge, and stress (Hafiz et al., 2023). So, university students, especially those who transitioned from home to a new place, full of different, unfamiliar eating options, encounter changes in dietary patterns and face greater challenges to follow the healthy dietary practices (Mishra and Anand, 2020).

Myanmar has many students who live and pursue higher education in many different countries around the world. According to the data of 2022, over 17,000 students from Myanmar were studying in foreign countries (ITA, 2023a, 2023b). Among the countries, Thailand is the most popular as it had approximately 3,700 of Myanmar students (ITA, 2023a, 2023b). So, the number in Thailand occupied about 22% of all Myanmar’s students in foreign countries. Chiang Mai Province, the second largest in Thailand, is one of the places that Myanmar students mostly choose for their studies in Thailand. According to the 2023 academic year data, 936 (27% of all international student population) studying in Chiang Mai Province were students from Myanmar (MHESI, 2023).

The non-medical factors that can affect the conditions of daily life and shape health outcomes are considered as the social determinants of health (WHO, 2024). According to the socio-ecological model of Bronfenbrenner, the social determinants are categorized into different levels, which are intrapersonal, interpersonal, organizational, and community levels. Although DD is important for individual nutritional requirements, many complex and multifactored reasons hinder people from eating healthy DD. According to (Osei-Kwasi et al., 2020), cognitive factors such as food preferences, mood, perceptions of diet quality and quantity, decreased self-esteem, and body satisfaction were associated with inadequate dietary diversity. The increased stress levels, low quality of life, time constraints, and physical inactivity contributed to improper eating behaviors (Osei-Kwasi et al., 2020). In (Alamirew et al., 2023), the religious norms such as fasting and food restrictions significantly impacted the status of minimum variety in the diet. Exploring the factors that foster or hinder DD at different levels is thus fundamental and crucial in promoting healthy behavior, lifestyle, and dietary intake. Understanding eating habits along with challenges will provide insights to identify unique needs and potential areas for intervention.

In a study, Myanmar students in Thailand reported that they encountered the challenges of not knowing what to order or how much it costs due to not having an English menu at universities’ cafeterias and they solved these problems by eating the same meals every day (Rhein and Jones, 2020). This has a very high possibility of seriously impacting the value and diversity of their daily diet, thereby affecting their nutritional status, health, and academic performance. This can lead to big negative consequences for both the origin and host countries. Firstly, the unhealthy and malnourished student populations probably lead to a burden on Thailand’s health system and impair the achievements of Thailand’s universities. Then, the poor dietary patterns of students become a blueprint for long-term dietary preferences and consumption patterns (Sprake et al., 2018). This will also affect the lifelong health outcomes and have the greatest impact on the public health of Myanmar when those students come back to the country. Investigating the DD of Myanmar students studying in Thailand is, therefore, a valuable and worthwhile area of study.

However, it is hard to find and almost nonexistent information on DD, influencing factors, and adaptation of Myanmar students to new dietary environments in Thailand on our current knowledge. This study, therefore, investigated dietary diversity of Myanmar university students in Chiang Mai province and identifies its social determinants at intrapersonal, interpersonal, organizational, and community levels of the socio-ecological framework.

MATERIALS AND METHODS

Study design, location and sample

A cross-sectional study was conducted from 3^rd October to 5^th November 2024 through a self-administered online survey with 170 Myanmar students studying at Chiang Mai University (CMU), the Rajamangala University of Technology Lanna (RMUTL), and Mahachulalongkornrajavidyalaya University (MCU) in Chiang Mai Province, Thailand.

Sample size determination and sampling technique

Sample size determination   

The sample size was calculated by estimating one population mean. Using standard deviation (SD) (σ)= 0.6 (Mediratta et al., 2023), the margin of error (d)= 10% (Mogeni and Ouma, 2022), and a 95% confidence interval (CI), the primary sample size was 138. After adjusting to a 20% non-response rate, 166 samples were finally required. However, as many samples as possible were collected, even exceeding the initial estimates, to enhance the generalizability and the external validity of the study (Sultana et al., 2022). After evaluating the samples according to the criteria, a total of 170 samples were deemed eligible for inclusion in the study.

Sampling technique

The samples were selected by the multistage non-probability sampling technique. Firstly, the three universities, namely Chaing Mai University (CMU), Rajamangala University of Technology Lanna (RMUTL), and Mahachulalongkornrajavidyalaya University (MCU), were purposively selected from a total of 8 universities in Chiang Mai Province. Then, a total of 203 responses were collected by convenient sampling techniques during the online survey. Finally, a total of 170 participants were selected for the study according to the inclusion criteria, exclusion criteria, and the completeness of their responses received. The inclusion criteria were aged 18-40 years, studying at any certificate, undergraduate, graduate, doctoral, or post-doctoral programs, and giving informed consent while having a food allergy was an exclusion.

Ethical considerations

This study was approved by the “Research Ethics Committee of the Faculty of Public Health, Chiang Mai University” with the reference number ET034/2024 on 20 September 2024.

Measurement tool

A survey questionnaire was used for this study. It contained three parts that are socio-demographic characteristics, dietary diversity assessment, and social determinant questions.

(a) Socio-demographic section consisted of a total of 17 questions to assess the information on age, gender, religion, university, educational status, marital status, family occupation, monthly family income, smoking status, alcohol consumption, length of stay in Chiang Mai, reported weight and height, difficulties due to language barrier, and history of underlying diseases.

(b) Dietary diversity was measured by the standardized questionnaire of the Food and Agriculture Organization of the United Nations (FAO) with the 24-hour recall method (FAO, 2018). There were 17 questions, 1. Cereals 2. White roots and tubers 3. Vitamin A-rich vegetables and tubers 4. Dark green leafy vegetables 5. Other vegetables 6. Vitamin A-rich fruits 7. Other fruits 8. Organ meats 9. Flesh meats 10. Eggs 11. Fish and seafood 12. Legumes, nuts, and seeds 13. Milk and milk products 14. Oils and fats 15. Sweets 16. Spices, condiments, and beverages 17. Eating outside the home (FAO, 2018). The respondents were instructed to administer the questionnaires according to FAO by recalling all foods and drinks eaten, including those consumed outside the home, last 24 hours, then filling in “Yes” or “No” in the food group list as per the separate ingredients (Horsey et al., 2019).

(c) Social Determinants: According to the adapted socio-ecological framework of this study, 36 questions were developed to evaluate the specific predetermined determinants at each stratum. At the intrapersonal level, 18 questions were asked for students’ knowledge, perception, eating three meals a day, food restrictions, snacking behaviors, skipping meals, eating way, stress, taste preference, food choice factors, financial constraints, monthly allowance, and self-efficacy. At the interpersonal level, there were eight questions to know eating companions, influencing persons who give dietary suggestions, suggested foods by influencing persons, importance of opinions of influencing persons, and influences of social media, social events, and celebrations. Also, the institutional-level availability of dietary advice, the existence of food policies and initiatives, and limitations to diverse food access were evaluated. Finally, seven questions were put to determine food accessibility, high food cost, cultural influences, religious restrictions, food advertisement, accessibility, and sources of dietary information at the community level.

The family income and monthly allowance for food were originally collected and analyzed in MMK and TBH respectively. But these were converted to the United States Dollars (USD) for presentation and reporting. For the conversion from the MMK to USD, the exchange rate (1USD= 2,100 MMK) of Central Bank of Myanmar during the study period was used. TBH was converted to USD according to (Suwanbamrung et al., 2024).

The content validity of the social determinant questions developed by the researcher was reviewed, commented, and scored by three experts, who are professors from Faculty of Public Health, with expertise in research methodology, behavioral science and public health administration. After revising the questions based on their feedback, the Item-Objective Congruence (IOC) of the questions was calculated, and each item got the IOC 1. Then, a try-out test was conducted with 30 samples that have the same characteristics as the study population to determine the reliability of the tool. When the internal consistency of the questionnaire was measured using SPSS software, the Cronbach's alpha coefficient was 0.934.

Data collection

The data was collected using the Google Forms questionnaire distributed to the participants via the student’s university email, in department group chat, and Myanmar student Facebook page. The informed consent was obtained from all participants electronically by clicking the box “agree to participate. To get the regular dietary intake of students by excluding the anomalies or special situations of the weekends and ensure that their 24-hour recall period was one of the usual weekdays, participants were requested to answer the questionnaires between Tuesday to Saturday. Also, it was sure by not gathering any response received on Sunday and Monday of the week.

Data analysis

For scoring of the DD, each food group in the list was valued as “0” for "No" which meant not consuming, or “1” for "Yes" which meant consuming food group in the last 24 hours (Zhao et al., 2020). The DDS was computed according to the FAO's guidelines by excluding fats and oils; discretionary foods; and spices, condiments, and beverages because these groups do not reflect the micronutrient adequacy of the diet (FAO, 2018). However, the frequencies of consuming these categories were evaluated since they could contribute to certain energy loads (FAO, 2018). Then, some food groups with similar nutritional properties were aggregated into the new forms of variables (FAO, 2018). Cereals (question no.1) and white roots and tubers (no.2) were combined into starchy staples; vitamin A-rich vegetables and tubers (no.3) and vitamin A-rich fruits (no.6) into other vitamin A-rich fruits and vegetables; other vegetables (no.5) and other fruits (no.7) into other fruits and vegetables, and flesh meats (no.9), and fish and seafood (no.11) into meat and fish (FAO, 2018). These new groups were 1 if one of the integrated categories was 1, but 0 if both were 0. So, a total of nine groups were contained in the DDS, and it was calculated by summing the number of food groups eaten out of 9 on the previous day (FAO, 2018).

Stata 15.1 was used for statistical analysis. The mean score was calculated to analyze the DDS (FAO, 2018). To summarize the data, descriptive analysis was performed with the frequencies (percentages) for categorical variables and mean (SD)/median (IQR) for continuous variables based on their distribution. To find out the determinants and the magnitude of the effect on the DDS, simple linear regression was done first by recruiting all socio-demographic and social determinant variables. Then, every significant factor was recruited in the initial model of multiple linear regression model to find out the combined effect of predictors controlling confounding factors. Then, the variables were removed one by one with backward elimination to get the final model. In eliminating the variables, the higher the P-value, the earlier the removal. Once a variable had been eliminated from the model, the likelihood ratio chi-square test statistic and the p-value were checked to ensure that the dropped variable did not significantly contribute to the model. Differences were finally deemed significant when the P-value was <0.05.

RESULTS

Socio-demographics of the participants

As presented in Table 1, 170 students participated in the study. 48.24% of them were females, with the median (IQR) age was 23±8 years. The majority, 92.35% were single, and 88.82% were Buddhist. Nearly half of the students (47.65%) were from CMU, and nearly three-quarters (73.53%) were undergraduates. Own business was mostly reported family occupation (37.06%), and the median (IQR) family income was 476.19 ± 952.38 USD. Regarding the biometric and lifestyle factors, the median (IQR) weight was 58 ± 17.50 kg, the median (IQR) height was 164.50 ± 13 cm, and the median (IQR) BMI computed from reported weight and height was 20.76 ± 4.99 kg/m². 138 students (81.18%) revealed themselves as non-smokers, while 73.53% described no current alcohol consumption, and 91.76% expressed no history of underlying disease. The median (IQR) length of stay in Chiang Mai Province was 9 ± 12months, and 42.94% revealed that they sometimes faced difficulties due to the language barrier.

Table 1. Socio-demographics of participants (n= 170).

Note: ¹Follow-up questions that are only applicable to certain participants based on their responses in the previous questions.

Food groups consumption and dietary diversity score

In evaluating the food group consumption from DD measurement using 24-hour recall, 86.47%, 35.29%, 58.24%, and 62.35% of students revealed eating cereals, white roots and tubers, vitamin A-rich vegetables and tubers, and dark green leafy vegetables, respectively. The frequency (%) of consuming other vegetables, vitamin A-rich fruits, and other fruits was 151 (88.82%), 81 (47.65%), and 85 (50.00%), respectively. 39.41% of participants consumed organ meat, 86.47% reported eating flesh meat, and 74.12% revealed having eggs. Fish and seafood; legumes, nuts and seeds; milk and milk products; oils and fats; sweets and spices, condiments, and beverages were eaten by 47.06%, 37.06%, 65.88%, 83.53%, 71.18%, and 79.41% of participants, respectively. 72.94% reported eating outside the home. The median (IQR) DDS was 6±3, with one score as the minimum and nine as the highest, as presented in Figure 1.

Figure 1. Food group consumption by participants (n=170).

Determinant factors of participants

Intrapersonal level

141 students (82.94%) had good DD knowledge. 62.35% perceived DD as frequently (often and always) important. For self-efficacy, only 28.24% were frequently (often and always) confident to eat DD every day. 89 students (52.35%) reported not eating three meals a day, and 154 out of 170 reported skipping meals with varying frequencies, while 16 never skipped. Among meal-skippers, breakfast was the most common skipped meal (n=102, 66.23%), and not readiness of food was the second highest reason for skipping meals (n=36, 23.38%) after time constraints (n=79, 51.30%). Two-fifths of 170 (40.00%) had food restrictions. Among them, 39.71% restricted fish and seafoods, and the main reason for food restrictions was dislike (n=41, 60.29%). Snacking behavior was found among 107 out of 170 students (62.94%), and “fruits/vegetables” was the most common type of snacks (n=53, 50.47%). Three-fifths of 170 (60.59%) reported cooking by self for meals. Half of the students (n=85) preferred other types of taste, such as arbitrary and versatile taste. The most common food choice factor was taste (n=111, 65.29%). Having financial constraints for food was revealed by 132 (77.65%), and the median monthly food allowance (IQR) was 100.95 ± 72.11 USD. A few proportion (11.18%) indicated that stress rarely impacted their eating practices.

Interpersonal level

Among 170 students, only a small percentage, 21.18% reported having meals with family members, and 9.41% with others (supervisors/advisors/professors and partners). 115 out of 170 received the dietary suggestions from influencing groups, varying across family, friends, and others. Among them, 70.43% revealed that they were encouraged to eat whole food or unprocessed foods, and 51.30% perceived that the opinions of influencing people were frequently (often, always) important to them. 165 out of 170 respondents joined the social events, varied across rarely, sometimes, often, and always, and 46.67% of them perceived that these celebrations sometimes influence their dietary behaviors. The majority (n=160) reported watching food blogs on social media at different frequencies, while 10 never watched, and nearly two-fifths (39.38%) of them expressed that they sometimes followed to eat like blogs.

Institutional/organizational level

Most of the students, (n=113, 66.47%) chose “don't know” on the availability of program/place for dietary advice at university, 112 (65.88%) also stated “don't know” whether the policies, initiatives, or programs to promote DD are existing, while nearly one-fourth (24.70%) expressed the presence of limitations for diverse food access at university.

Community level

Nearly four-fifths of 170 participants (79.41%) reported that they had easy access to healthy foods, while just over three-fifths (62.35%) revealed that healthy foods are expensive in their area. 104 (61.18%) expressed no cultural influence on their DD, while a few proportions (16.47%) described that there are religious restrictions on their eating habits. Among the food advertisement channels, online social media was the highest chosen platform (n=138, 81.18%), while billboard followed as third (n=19, 11.18%). Regarding the accessibility to dietary information in the area, slightly more than half of the students (n=96, 56.47%) responded “Yes”, and the majority of those (83.33%) presented that they received the dietary information from online social media. Please see the summary of determinant factors at each level of socio-ecological framework in Table 2 below and details in Table S1.

Table 2. Summary of determinant factors among participants. (n= 170).

Note:   ¹Follow-up questions that are only applicable to certain participants based on their responses in the previous questions. ²Multiple-choice questions that allow participants to choose more than one option.

Finding social determinants influencing dietary diversity score

Simple linear regression and initial model of multiple linear regression

In the simple linear regression to identify potential predictors associated with DDS, a total of 36 variables, 3 socio-demographic factors, 17 intrapersonal, 6 interpersonal, 3 institutional, and seven community-level variables were statistically significant. When these variables were put in the initial model of multiple linear regression to find out the combined effect of predictors controlling confounding factors, only two out of 36 variables, namely “skipping meals due to not readiness of foods” (coefficient= -0.784, 95% CI= -1.514 to -0.054 and P-value= 0.036) and “food choice by taste” (coefficient= 0.664, 95%CI= 0.008 to 1.319 and P-value= 0.047) were significant. Please see the summary of simple linear regression and initial model of multiple linear regression in Table 3 below and details in Table S2.

Table 3. Summary of simple linear regression and initial model of multiple linear regression for social determinants with dietary diversity score.

Note: ¹Initial model of multiple linear regression.  *Significant with P-value < 0.05.

Final model of multiple linear regression

After backward elimination, 13 variables were maintained in the final model of multiple linear regression as presented in Table 4. Among them, two demographic factors, gender and religion, were associated with DDS as females had a lower DDS of 0.553 scores than males (95% CI= -1.075 to -0.031, P-value= 0.038) and other religions also showed a decrease in DDS, 0.964 scores compared to Buddhists (95% CI= 1.745 to -0.183, P-value= 0.016). At the intrapersonal level, self-efficacy, students who frequently trusted their ability to eat diverse diets were likely to have a higher DDS of 0.976 scores (95% CI= 0.211 to 1.741, P-value= 0.013) than those who infrequently did. Participants who skipped meals due to “not readiness of food” tended to have a lower DDS of 0.926 scores (95% CI= -1.554 to -0.299, P-value= 0.004) compared to their counterparts. Respondents who perceived stress rarely impacted their eating meals showed a decrease in DDS, 1.387 scores (95% CI= -2.631 to -0.143, P-value= 0.029) than those who perceived never.

At the interpersonal level, students who usually had meals with family members tended to have a higher DDS by 0.989 score (95% CI= 0.321 to 1.656, P-value= 0.004), and those who had meals with other people (supervisors/advisors/professors, partners) also showed an increase in DDS, 1.275 scores (95% CI= 0.336 to 2.214, P-value= 0.008) than their counterparts. Moreover, respondents who received dietary suggestions from friends were likely to have a higher DDS of 0.754 scores (95% CI= 0.095 to 1.414, P-value of 0.025) compared to those who reported no one. At the organizational level, students who described having limitations for diverse food access at university significantly had a higher DDS of 0.813 scores (95% CI= 0.172 to 1.454, P-value= 0.013) than those who did not perceive it. As the community level, it was found that participants reporting religious restrictions on their eating habits were likely to have an increase in DDS, 1.065 scores (95% CI= 0.335 to 1.796, P-value= 0.005), and students who indicated advertising food on billboards in their environment had a lower DDS of 0.928 scores (95% CI= -1.728 to -0.128, P-value= 0.023) compared to their counterparts.

Table 4. Final model of multiple linear regression for social determinants with dietary diversity score.

DISCUSSION

Although several studies have identified the correlates of dietary diversity, very few considers the broader social and demographic influences and almost none was conducted among Myanma students in Thailand. This study extends the evidence by examining these relationships within the socioecological framework in the context of Myanmar university students in Chiang Mai Province. Gender, self-efficacy and stress at the intrapersonal level, support from peers and family at the interpersonal layer, organizational limitations to diverse food access, and religious restrictions and food advertisement channels at the community level were influencing the DDS of university students.

Dietary diversity score

The mean DDS 6.21 ± 1.99 out of 9 seems moderately good, and most participants have relatively high values, whereas a few scores are extremely low. This result was contrary to the study in rural Ghana, which found 84.7% of the participants had poor DD with a mean score of 3.8 ± 0.8 (Wiafe et al., 2023). In another study in Bangladesh, the DD status of university students was indicated enough among 55.8% while the other 44.2% had inadequate DD (Moon et al., 2024). Also, a study at the University of Dhaka expressed that DD of approximately 63% of participants was adequate, but 37% did not meet the minimum requirement (Sultana et al., 2019). Another study in southern Nigeria explored that the DDS of 392 students (49.0%) was high, but it was insufficient in 26.5% (Omage and Omuemu, 2018). Having reasonable DDS among the students in this study could be due to the study location being within the urban area with easily available and abundant food options, proven by a large proportion of students reporting food accessibility in their area. In contrast, overestimating the variety of food categories they had or giving the desirable answers with social perspectives is one of the considerable facts in the self-reported assessment. 

Determinants influencing on the dietary diversity score of students

Sociodemographic factors

The result that female students had lower DDS than males is against the study among the students in Syria, which did not find any gender difference in DD (Younes, 2024). This may be due to the differences in sample characteristics, statistical analysis, and environmental and/or cultural contexts between studies. In contrast, the finding was supported by a study in southern Nigeria (Omage and Omuemu, 2018), in which DDS was significantly higher in males than females. However, the effect of gender on DDS cannot be justified well with the insights of the current result. This could be due to the biological variation forcing males to require higher energy and eat more diverse foods, or men having more food exposure by doing outdoor activities or wandering around the city, or many other factors. So, further research in different settings would be necessary to strengthen this evidence and generalizability.

Religion was found to predict the DDS, as other religions had lower DDS than Buddhists. In Tanzania, a higher DDS was explored among Muslims compared to Christians among university students in 2024 (Mgetta and Muhimbula, 2024). Another study also reported that Muslim students had 70% lower odds of inadequate DD than Orthodox Christians (Gonete et al., 2017). So, these comparable results among different religious groups in various backgrounds consistently support the reliability of our findings. In contrast, a significant association between religion and DD was not found in (Endalifer et al., 2021). Nevertheless, we acknowledge the explanation in (Mgetta and Muhimbula, 2024), because the uneven distribution of Buddhist students, contributing to nearly 90% of the entire sample size, may be attributed to the significant impact of religion on DDS in this study.

In this study, an association could not be explored between the age of the students and the dietary diversity score. This was against the study conducted among students in Khulna University because samples who were in the age between 21 to 25 years old had higher dietary diversity than students who were 17–20 years old (AOR = 1.79, 95% CI = 1.04–3.08, P = 0.035) in that study (Moon et al., 2024). Also, a study at a private university in southern Nigeria found that the age categories of the undergraduate students were significantly associated with the dietary diversity score at P < 0.001(Omage and Omuemu, 2018). These differences may be attributed to variations in the study population, geographical context, and methodological approaches. Additionally, differences in measurement tools and analytical strategies across studies may have contributed to the observed discrepancies.

In the current study, family income was not a statistically significant social determinant of the dietary diversity of the students. Research carried out in Gurage Zone, Southwest Ethiopia, identified that the monthly family income was significantly associated with DDS (AOR = 1.56, 95% CI = 1.28-1.9, P = 0.001) (Worku et al., 2017). In the urban context of Chiang Mai Province, a variety of food options are accessible at an affordable price for the student’s budget. The economic capacity of the students, including the family income and the monthly food allowance, does not therefore stand as the predictor of dietary diversity status in this study in the existence of other crucial determinants. Another possibility is that the targeted student population of this research is better aware of good nutrition and a balanced diet than the studied group of most previous literature, and it may have some impact in reducing the financial barrier to dietary diversity. Also, the social influence among the peers seems to overwhelm the affordability issues in this meal practice. Anyhow, this finding may reflect evolving socioeconomic conditions compared with earlier studies.

Intrapersonal level determinants

Self-efficacy, frequently (often and always) confident to eat diverse diet, determined to have a higher DDS. This is against a study among rural adolescents in Pakistan that described no association between self-efficacy and DDS (Baxter et al., 2022). The empowerment of rural Pakistani girls has cultural boundaries limiting the chance of raising preferences, which then affects food intake (Baxter et al., 2022). This could reflect the differing results between the two studies. However, the finding is comparable to the insights of a study in Bangladesh that indicated self-efficacy as one of the significant predictors of maternal DD (Nguyen et al., 2017). Nonetheless, this finding reveals that programs improving ability of students for healthy dietary behaviors, for instance, dietary counseling, cooking demonstration, or culinary education, should be initiated in universities as people with high self-efficacy can use their capabilities to maintain the behavior even with the challenges (Suriyawong and Pipatpiboon, 2022).

To the best of our knowledge, studies on the association between skipping meals due to food not being ready and the DDS are almost nonexistent. This study found that participants with this practice had lower DDS than their counterparts. The literature in Bangladesh explored that students who solved starvation by eating foods available/cooked by themselves had higher DDS than those who stayed hungry without eating (Moon et al., 2024). This supports our finding, regardless of the varied statistical analysis between research. Moreover, this association could be interpreted through the lens of the socio-ecological framework which emphasizes that the countless individual-level predictors can push autonomous students into unhealthy practices and lifestyles in a predictable way. However, further investigation is recommended to strengthen this interpretation.

In this study, respondents who perceived stress rarely impacted eating meals had lower DDS than those who never perceived it. In Pakistan, the association between stress-like feelings and DDS was not found among adolescent girls (Baxter et al., 2022). In contrast, the finding is supported by (Kabir et al., 2018) because it found that stress was a major predictor of the dietary practices of the students. Nevertheless, student life certainly imposes stress for countless reasons, and it tends to highly affect the student’s diet. This is proven by a large proportion of respondents reported that stress impacts eating meals varied across the frequencies in this study. So, academic institutions should be the platform to promote psychosocial support for a better stress coping mechanism of students.

Intrapersonal-level determinants

The social factors, eating meals with family members, and others (professors/advisors/supervisors, seniors, juniors, and partners) increased the DDS of students in our study. A study in Woldia did not find any correlation between the adolescent DD and eating partners (Endalifer et al., 2021). However, this finding corresponds to the insights of a qualitative study in a public university in Bangladesh that described how a few students preparing meals together help them to make a variety of meals with reduced cost, labor, and time, and so they avoid monotonous dishes (Kabir et al., 2018). Even if the food from on-campus canteens is eaten, students tend to have several dishes as the costs are shared between them (Kabir et al., 2018). Anyhow, it seems very challenging for students to eat meals with other people, as given more than half of students in this survey responded having meals alone. So, creative actions that would encourage group-based eating behavior are worthwhile to promote in the universities and/or students’ residences.

In the present study, students who received dietary suggestions from friends had higher DDS compared to those with no support from anyone. This finding supports the social cognitive theory which explained that supportive social relationships encourage people to sustain healthy behaviors (Suriyawong and Pipatpiboon, 2022). In a mixed-method research among Ulster University students, eating behaviors were positively or negatively impacted by peers (Hafiz et al., 2023). Also, a qualitative study in Bangladesh revealed that friends in the same residents/rooms/classes were influencing dietary choices (Kabir et al., 2018). Beyond the student population, DD was 2.30 (95% CI [1.33–03.98]) lower in the migrant population with no support from friends/organizations than their counterparts (Essayagh et al., 2024). Exploring the significant role of peers on the students' DDS helps us identify the potential target groups and understand how to effectively establish the institutional behavior-change programs.

Institutional/organizational-level determinants

Contrary to common assumptions, students who revealed limitations in accessing diverse foods at university tend to have a higher DDS than those who reported no limitations. A study also revealed that the scarcity of healthy foods at the university was the key predictor of optimal consuming practices (Hilger et al., 2017). Even though studies directly addressing this relationship are limited, it can be understood by the Health Belief Model, in which individuals may adopt positive behaviors even with barriers if they have sufficient self-efficacy and perceive benefits. It is proved by a large proportion of participants in this study because over 90% of the students perceived that DD is important for better health outcomes and showed confidence to eat DD, although the frequency differed. Promoting healthy and diverse foods in institutional settings is thus worth improving the dietary intake and nutrition of students.

For the other two institutional determinants, students, expressing the availability of dietary advices and the existence of food policies at their universities, had a higher dietary diversity score in the simple linear regression analysis. However, these associations did not remain statistically significant in the multiple linear regression model after adjustment for other variables. The observed associations in the unadjusted analysis may be influenced by other factors included in the adjusted model. From a programmatic perspective, the institutional dietary education programs and supportive policies may function as part of a broader organizational environment and comprehensive approach to optimum dietary diversity of students. Although these variables were not statistically significant in the adjusted analysis, the positive direction of associations indicates a potential role in promoting healthy eating behaviors. Due to the limited evidence in this context, further studies in other methodological approaches with larger sample sizes are demanding to better understand the influence of the dietary advisory programs and policies/initiatives in the institutional settings.

The assessment of institutional determinants in this study was limited only to the perceived availability of dietary advice, the presence of food policies, and limitations in accessing diverse foods. However, other organizational factors, such as food storage facilities, types, quantity, and quality of food outlets located in and around the universities, may also play an important role in shaping the dietary behaviors of the students. Since all the study locations are in the central part of Chaing Mai and numerous street food vendors, one-stop marts, night food markets, automated vending machines, and cafeterias exist in and around the campuses, library, bus stops, community centers, public areas and near the student’s dormitories, the inspection, quality control, and supportive action to the surrounding food retails are all worthwhile for better dietary intake. These factors were not assessed in the present study. So, future researchers are suggested to consider a more comprehensive assessment of the food environments in universities to better understand their influence on dietary behaviors.

Community-level determinants

A higher DDS was found among the participants who indicated religious restrictions on eating habits in the present study. It corresponds to the insights of a study which found that religious beliefs and traditions influenced the dietary practices of numerous students (Hafiz et al., 2023). This supports the socio-ecological model which explained that factors at the environmental level, including the cultural norms, and religious influences play a major role in determining behaviors in various ways. So, public administrators and/or leaders should enable the initiatives sensitive and strategic, giving more consideration to traditional, cultural, and religious factors. Another significant factor at the environmental level, advertising food on billboards, appeared to be linked to the lower DDS of the students. This is contrary to a study in Ethiopia that observed no association between advertising food on public displays and the dietary scores (Trübswasser et al., 2022). Although limiting the literature on this specific association, this current finding reflects the unique environmental determinant on the DD of students and suggests a potential link that necessitates further exploration. So, future researchers are recommended to emphasize the impact of mass communication on dietary behavior and diversity across settings. Also, the public and private stakeholders should ensure the public messaging does not mismatch the target audience and communicate with the people effectively.

LIMITATIONS

The cross-sectional study design limits the ability to deduce causal relationships, and self-reported data may introduce biases in behavior reporting. Also, the convenient sampling method may limit the generalizability to the wider population. So, future studies that would address these limitations are suggested to conduct for better understanding of the multi-level determinants on the dietary diversity among the university students.

CONCLUSION

This study identified several factors influencing dietary diversity across different levels of the adapted socioecological model. Gender, intrapersonal self-efficacy and stress, interpersonal support from peers and family, organizational limitations to diverse food access, and community-level religious restrictions and advertisement channels were influencing the dietary diversity score of university students. Comprehensives underscore the importance of a comprehensive multi-level approach in dietary behavior change interventions. By addressing barriers and leveraging facilitators at each level, further initiatives can more effectively promote dietary diversity of the students.

RECOMMENDATION

To practically promote healthy eating practices and improve the dietary diversity of the students, comprehensive and inclusive interventions rather than focusing on awareness- raising are important. Universities and public health authorities should develop healthier food options in campus cafeterias, peer-oriented nutritional behavior-change promotion, and psychosocial support program for better coping mechanisms of stress to encourage students to follow proper dietary practices. All these measures should be gender-specific and culturally or traditionally appropriate to enhance their effectiveness.

AUTHOR CONTRIBUTION

Phyoe Phyoe NiLar Kyi: Conceptualization (Lead), Methodology (Equal), Validation (Equal), Formal analysis (Equal), Investigation (Lead), Writing - Original Draft (Lead), Visualization (Lead), Project administration (Lead); Jukkrit Wungrath: Conceptualization (Lead), Methodology (Equal), Formal analysis (Equal), Writing - Review & Editing (Lead), Supervision (Lead); Pallop Siewchaisakul: Conceptualization (Supporting), Methodology (Equal), Formal analysis (Equal), Writing - Review & Editing (Equal), Visualization (Supporting), Supervision (Equal).

REFERENCES

Alamirew, S.K., Lemke, S., Stadlmayr, B., and Freyer, B. 2023. Dietary behaviour and sociocultural determinants of dietary diversity among rural women of reproductive age: A case of Amhara Region, Ethiopia. Nutrients. 15: 3369. https://doi.org/10.3390/nu15153369

Baxter, J.-A.B., Wasan, Y., Islam, M., Cousens, S., Soofi, S.B., Ahmed, I., Sellen, D.W., and Bhutta, Z.A. 2022. Dietary diversity and social determinants of nutrition among late adolescent girls in rural Pakistan. Maternal & Child Nutrition. 18: e13265. https://doi.org/10.1111/mcn.13265

Endalifer, M.L., Andargie, G., Mohammed, B., and Endalifer, B.L. 2021. Factors associated with dietary diversity among adolescents in Woldia, Northeast Ethiopia. BMC Nutrition. 7(1): 27. https://doi.org/10.1186/s40795-021-00430-6

Essayagh, F., Essayagh, M., Lambaki, A., Naji, A., Essayagh, S., and Essayagh, T. 2024. Determinants associated with low dietary diversity among migrants to Morocco: A cross sectional study. Scientifc Reports. 14: 8361. https://doi.org/10.1038/s41598-024-59082-8

FAO. 2018. Dietary assessment: A resource guide to method selection and application in low resource settings. Food and Agriculture Organization of the United Nations, Rome.

Gonete, K.A., Tariku, A., Wami, S.D., and Akalu, T.Y. 2017. Dietary diversity practice and associated factors among adolescent girls in Dembia district, northwest Ethiopia, 2017. Public Health Reviews. 41: 23. https://doi.org/10.1186/s40985-020-00137-2

Hafiz, A.A., Gallagher, A.M., Devine, L., and Hill, A.J. 2023. University student practices and perceptions on eating behaviours whilst living away from home. International Journal of Educational Research. 117: 102133. https://doi.org/10.1016/j.ijer.2022.102133

Hilger, J., Loerbroks, A., and Diehl, K. 2017. Eating behaviour of university students in Germany: Dietary intake, barriers to healthy eating and changes in eating behaviour since the time of matriculation. Appetite. 109: 100-107. https://doi.org/10.1016/j.appet.2016.11.016

Horsey, B., Swanepoel, L., Underhill, S., Aliakbari, J., and Burkhart, S. 2019. Dietary diversity of an adult Solomon Islands population. Nutrients. 11: 1622. https://doi.org/10.3390/nu11071622

ITA. 2023a. Burma- education and training sector snapshot. International Trade Administration.

ITA. 2023b. Burma Study Abroad. International Trade Administration.

Kabir, A., Miah, S., and Islam, A. 2018. Factors influencing eating behavior and dietary intake among resident students in a public university in Bangladesh: A qualitative study. PLoS One. 13(6): e0198801. https://doi.org/10.1371/journal.pone.0198801

Martin-Prevel, Y., Arimond, M., Allemand, P., Wiesmann, D., Ballard, T.J., Deitchler, M., Dop, M., Kennedy, G., Lartey, A., Lee, W.T., et al. 2017. Development of a dichotomous indicator for population-level assessment of dietary diversity in women of reproductive age. Current Developments in Nutrition. 1: e001701. https://doi.org/10.3945/cdn.117.001701

Mediratta, S., Ghosh, S., and Mathur, P. 2023. Intake of ultra-processed food, dietary diversity and the risk of nutritional inadequacy among adults in India. Public Health Nutrition. 26(12): 2849-2858. https://doi.org/10.1017/S1368980023002112

Merga, G., Mideksa, S., Dida, N., and Kennedy, G. 2022. Dietary diversity and associated factors among women of reproductive age in Jeldu district, West Shoa Zone, Oromia Ethiopia. PLoS One. 17(12): e0279223. https://doi.org/10.1371/journal.pone.0279223

Mgetta, N.J. and Muhimbula, H. 2024. Dietary diversity and anthropometric status of students at universities in Dodoma and Morogoro Regions, Tanzania. Tanzania Journal of Health Research. 25: 2. https://doi.org/10.4314/thrb.v25i2.14

MHESI. 2023. Student Information for Academic Year 2023. Ministry of Higher Education, Science, Research & Innovation.

Mishra, S. and Anand, S. 2020. Migration and dietary diversity changes among the students: Case study of the university of Delhi in India. Space and Culture. 8(3): 58-70. https://doi.org/10.20896/saci.vi0.906

Mogeni, B.K. and Ouma, L.O. 2022. Dietary patterns, behaviours, and their associated factors among university students in coastal Kenya. Cogent Food and Agriculture. 8(1): 2132873. https://doi.org/10.1080/23311932.2022.2132873

Moon, M.I., Islam, T., Tohan, M.M., Shah, B.R., Juhi, N.J., Anni, T.T., Islam, A.M.M., Akter, K.M., Howlader, M.H., and Rahman, M.A. 2024. Dietary diversity among students of Khulna University, Bangladesh. Dr. Sulaiman Al Habib Medical Journal. 6: 92‑104. https://doi.org/10.4103/DSHMJ.DSHMJ_32_24

Mukherjee, A., Paul, S., Saha, I., Som, T.K., and Ghose, G. 2018. Dietary diversity and its determinants: A community‑based study among adult population of Durgapur, West Bengal. Medical Journal of Dr. DY Patil Vidyapeeth. 11(4): 296-301. https://doi.org/10.4103/mjdrdypu.mjdrdypu_15_18

Nguyen, P.H., Sanghvi, T., Kim, S.S., Tran, L.M., Afsana, K., Mahmud, Z., Aktar, B., and Menon, P. 2017. Factors influencing maternal nutrition practices in a large scale maternal, newborn and child health program in Bangladesh. PLoS One. 12(7): e0179873. https://doi.org/10.1371/journal.pone.0179873

Nithya, D.J. and Bhavani, R.V. 2017. Dietary diversity and its relationship with nutritional status among adolescents and adults in rural India. Journal of Biosocial Science. 50: 397-413. https://doi.org/10.1017/S0021932017000463

Omage, K. and Omuemu, V.O. 2018. Assessment of dietary pattern and nutritional status of undergraduate students in a private university in southern Nigeria. Food Science & Nutrition. 6: 890-1897. https://doi.org/10.1002/fsn3.759

Osei-Kwasi, H., Mohindra, A., Booth, A., Laar, A., Wanjohi, M., Graham, F., Pradeilles, R., Cohen, E., and Holdsworth, M. 2020. Factors influencing dietary behaviours in urban food environments in Africa: A systematic mapping review. Public Health Nutrition. 23(14): 2584-2601. https://doi.org/10.1017/S1368980019005305

Rhein, D. and Jones, W. 2020. The impact of ethnicity on the sociocultural adjustment of international students in Thai higher education. Educational Research for Policy and Practice. 19: 363-388. https://doi.org/10.1007/s10671-020-09263-9

Sprake, E.F., Russell, J.M., Cecil, J.E., Cooper, R.J., Grabowski, P., Pourshahidi, L.K., and Barker, M.E. 2018. Dietary patterns of university students in the UK: A cross-sectional study. Nutrition Journal. 17(1): 90. https://doi.org/10.1186/s12937-018-0398-y

SSP. 2024. The state of food security and nutrition in Myanmar 2022-23: findings from six rounds of the Myanmar household welfare survey. Strategy support program working paper 45. International Food Policy Research Institute. Washington, DC.

Sultana, M., Hasan, T., and Shaheen, N. 2019. Dietary diversity and nutritional status of female residential students in University of Dhaka, Bangladesh. Indian Journal of Public Health Research & Development. 10(6): 644. https://doi.org/10.5958/0976-5506.2019.01349.4

Sultana, M.S., Islam, M.S., Sayeed, A., Koly, K.N., Baker, K., Hossain, R., Ahmed, S., Ferdous, M.Z., Mubarak, M., Potenza, M.N., et al. 2022. Food addiction, orthorexia nervosa and dietary diversity among Bangladeshi university students: A large online survey during the COVID-19 pandemic. Journal of Eating Disorders. 10: 163. https://doi.org/10.1186/s40337-022-00680-0

Suriyawong, W. and Pipatpiboon, N. 2022. Social cognitive theory-based interventions on healthy lifestyles for hypertensive patients: A systematic review. Chiang Mai University Journal of Natural Sciences. 21(3): e2022040. https://doi.org/10.12982/CMUJNS.2022.040

Suwanbamrung, C., Srinam, B., Promkool, P., Suwannakarn, W., Siripanich, S., Rahman, M.S., and Stanikzai, M.H. 2024. Uptake of COVID-19 vaccine among high-risk urban populations in Southern Thailand using the COM-Bmodel. PLoS One. 19(3): e0300509. https://doi.org/10.1371/journal.pone.0300509

Trübswasser, U., Talsma, E.F., Ekubay, S., Poelman, M.P., Holdsworth, M., Feskens, E.J.M., and Baye, K. 2022. Factors influencing adolescents' dietary behaviors in the school and home environment in Addis Ababa, Ethiopia. Frontiers in Public Health. 10: 861463. https://doi.org/10.3389/fpubh.2022.861463

WHO. 2024. Social Determinants of Health. World Health Organization.

Wiafe, M.A., Apprey, C., and Annan, R.A. 2023. Dietary diversity and nutritional status of adolescents in rural Ghana. Nutrition and Metabolic Insights. 16: 11786388231158487. https://doi.org/10.1177/11786388231158487

Worku, M., Hailemicael, G., and Wondmu, A. 2017. Dietary diversity score and associated factors among high school adolescent girls in Gurage zone, southwest Ethiopia. World Journal of Nutrition and Health. 5(2): 41-45. https://doi.org/10.12691/jnh-5-2-3

Younes, S. 2024. Gender difference in nutritional knowledge, dietary pattern, and nutritional status of undergraduates in Tartous University, Syria. Journal of Biomed Research. 5(1): 8-16. https://doi.org/10.46439/biomedres.5.038

Zhao, A., Li, Z., Ke, Y., Huo, S., Ma, Y., Zhang, Y., Zhang, J., and Ren, Z. 2020. Dietary diversity among Chinese residents during the COVID-19 outbreak and its associated factors. Nutrients. 12: 1699. https://doi.org/10.3390/nu12061699

Zhou, Y., Wang, J., Duan, Y., Luo, X., Wan, Z., Luo, Y., Li, Y., Wang, Y., and Xie, J. 2022. Dietary diversity and determinants of young adults in central China: A cross-sectional study from 2015 to 2020. Frontiers in Nutrition. 9: 931107. https://doi.org/10.3389/fnut.2022.931107

OPEN access freely available online

Natural and Life Sciences Communications

Chiang Mai University, Thailand. https://cmuj.cmu.ac.th