Fear or security? The role of nutrition literacy on sustainable and balanced dietary adherence

Abstract

Objective

The aim of this study is to examine the impact of food security and scarcity perception on sustainable and balanced dietary preferences among university students and to evaluate the regulatory role of nutrition literacy in these relationships.

Method

A total of 1200 female students aged 19–30, studying at state universities in Türkiye, were included in the study conducted with a cross-sectional design. Data were collected through face-to-face surveys; the Eating Information and Nutrition Literacy Assessment (EINLA), Adult Food Security Survey Module (AFSSM), Anticipated Food Scarcity Scale (AFSS), and Nutrition Based Eat Index (NB_EAT) were used to measure the related constructs. The analyzes were conducted using hierarchical regression and moderator effect tests.

Results

The mean NB_EAT score of the participants was 4.19 ± 1.04, and 64.8% were at a low level of adaptation. The food security score was found to be 23.14 ± 7.66, the hunger perception score was 43.34 ± 13.31, and the EINLA score was 29.68 ± 2.94. In regression analyzes, food insecurity is a significant predictor in terms of sustainable preferences (β = 0.097, p = 0.001). The perception of malnutrition was a positive determinant (β = 0.106, p = 0.004). Perceived scarcity was not found to be a significant predictor (p > 0.05). Nutrition literacy alone did not make a significant contribution, but in interaction with food security, it had a moderating effect on sustainable choices. In subgroup analyzes, a positive relationship was found between food security and sustainable preferences in individuals with sufficient nutrition literacy, while the relationship was not significant in those with insufficient literacy.

Conclusion

The findings indicate that food security is a stronger determinant of sustainable dietary choices than the perception of scarcity. While nutrition literacy is not effective in influencing perceptions of scarcity, it plays a role in strengthening the relationship between food security and sustainable choices. The study reveals that policies supporting sustainable nutrition should not be limited to structural improvements alone, but should also be supported by interventions aimed at improving individuals’ nutritional literacy. As this study was conducted exclusively among female university students, the findings should be interpreted within this demographic context.

Introductıon

Global food systems face multiple crises, including climate change, depletion of natural resources, loss of biodiversity, and increasing food insecurity. In this context, individuals adopting sustainable eating habits has become a critical necessity for both environmental sustainability and public health [1]. Sustainable nutrition is directly related not only to individual health but also to ecological and socioeconomic systems. Individuals’ sustainable eating habits are influenced by multidimensional structural and cognitive factors such as their level of environmental awareness, socioeconomic conditions, access to food, and level of knowledge about nutrition [2, 3, 4].

Among these factors, food security and scarcity perception stand out as key determinants guiding individuals’ consumption decisions. Scarcity perception is a subjective belief system that reflects an individual’s sense of inadequacy regarding current or future resources. This belief can lead individuals to engage in short-term and impulsive consumption behaviors [5, 6]. In contrast, food security reflects a more objective situation, meaning continuous, safe, and adequate access to food [7, 8]. Both structures influence an individual’s consumption habits. However, the effects of these structures on sustainable nutrition behaviors have not yet been sufficiently addressed in a comprehensive manner.

Nutrition literacy refers to an individual’s capacity to acquire, evaluate, interpret, and apply food-related knowledge to their daily life [9, 10]. This competency can facilitate individuals’ orientation toward sustainable choices and support them in making more rational and long-term decisions in the face of stressors such as scarcity perception or food insecurity. Therefore, nutrition literacy may be a cognitive protective or regulatory factor on sustainable and balanced nutrition choices.

In the current literature, nutrition literacy, food security, and sustainable and balanced eating behaviors are generally examined under separate headings; this structure is either dual (nutrition literacy and food security) [11–14], or tripartite [15, 16]. Perceived scarcity, on the other hand, has not yet been sufficiently addressed in research due to its recent emergence as a concept [5, 6]. Efforts to address these gaps will contribute to the more effective planning of awareness and education programs aimed at supporting individuals’ sustainable and balanced eating habits. In this regard, the aim of this study is to evaluate the effect of nutrition literacy, food security, and scarcity perception on sustainable and balanced nutrition preferences using a comprehensive model and to examine the regulatory role of nutrition literacy in these relationships. The primary objective of this study is to examine the role of food security and scarcity perception in individuals’ sustainable and balanced dietary preferences. The research questions explore whether these two structural variables significantly predict dietary preferences, whether their effects differ according to nutrition literacy, and whether nutrition literacy plays a regulatory role.

Materıal and methods

Study design

This research was conducted within the framework of a cross-sectional survey model based on quantitative research methods. Data on individuals’ food consumption was evaluated as sustainable nutrition preferences, and the aim was to examine the structural and cognitive factors affecting these preferences. The necessary ethical approval for the research was obtained from the Gazi University Ethics Committee. The study was conducted in accordance with the ethical principles of the Helsinki Declaration, and written consent was obtained from all participants.

Sample characterisation and recruitment

The study sample consisted of female university students aged 19–30 years, enrolled in state universities located in Ankara, Türkiye. Participants were recruited face-to-face through campus-based announcements, classroom visits, and notices placed in dormitories and common campus areas. Prior to data collection, students were briefly informed about the purpose of the study, and those who volunteered were included. Participation was entirely voluntary, and no incentives were provided.

A total of 1,263 students were approached in person, and after excluding five male respondents and 58 participants whose questionnaires were incomplete or could not be verified, the final analytical sample included 1,200 participants.

Inclusion criteria were being female, aged 19–30 years, enrolled in a public university, and not studying in the Department of Nutrition and Dietetics. Exclusion criteria included male gender, age outside the specified range, enrollment in private/foundation universities, and majoring in Nutrition and Dietetics.

Female participants were specifically chosen to ensure a homogeneous sample and minimize gender-related variability in nutrition literacy and eating behavior. Additionally, prior field observations indicated that female students tend to respond more carefully and consistently to self-administered questionnaires, thereby improving the reliability of the data collected.

An a priori power analysis was performed using G*Power version 3.1.9.7 for multiple linear regression (fixed model, R² deviation from zero). Assuming a small-to-moderate effect size (f² = 0.05), α = 0.05, and desired power (1–β) = 0.90, the minimum required sample size was calculated as 1,239 participants. The final sample of 1,200 participants provided adequate power to detect small-to-moderate effects across the regression models.

Data collection tools

The questionnaire form prepared for data collection was administered using the face-to-face interview method and took approximately 25–30 min to complete. The questionnaire consists of six sections:

General information

The first section of the questionnaire used in the study included participants’ introductory information. In this context, basic sociodemographic characteristics such as participants’ age, university and department of study, academic period, marital status, employment status and occupation, and type of housing were collected. In addition, the educational level of the parents, the monthly income of the family and the individual, the amount of expenditure allocated for nutrition, and income adequacy were inquired. In terms of health status, the presence and type of chronic diseases, history of dieting and types of diets applied, physical activity habits, and types of sports practiced were evaluated. Participants’ knowledge or education regarding nutrition, the source of this knowledge, and trusted sources of information were also determined. Finally, anthropometric data such as height, weight, and body mass index (BMI) were obtained.

Eating habits

Participants were asked to indicate the number of main and snack meals they consumed per day, and whether they skipped any main meals and, if so, which meal they skipped and why. Reasons for skipping meals included lack of habit, lack of appetite, waking up late, desire to lose weight, lack of time, and other individual reasons.

Adult Nutrition Literacy Assessment Tool (EINLA)

The Evaluation Instrument of Nutrition Literacy on Adults (EINLA), developed and validated by Cesur et al. [17], was used to assess participants’ nutrition literacy levels. The instrument was originally designed for Turkish adults aged 18–64 and has demonstrated strong psychometric properties (Cronbach’s α = 0.75,test–retest r = 0.85).

The scale consists of 35 items distributed across five domains: (1) general nutrition knowledge, (2) reading comprehension and interpretation, (3) food group knowledge, (4) portion sizes, and (5) food label reading and numerical literacy. Each item has a single correct response coded as 1 (correct) and 0 (incorrect or blank), with a total score range of 0–35.

Nutrition literacy levels are classified as inadequate (0–11 points), borderline (12–23 points), or adequate (24–35 points). The instrument’s content validity was confirmed through expert review (n = 10), and construct validity was established via factor analysis (KMO = 0.76; Bartlett’s χ², p < 0.001). The tool was designed specifically for the Turkish population and is widely used in community and university settings to assess functional, interactive, and critical nutrition literacy skills.

In this study, the validated scoring and classification criteria proposed by Cesur et al. [17] were adopted without modification. Internal reliability in the current sample was satisfactory (Cronbach’s α = 0.88) [17].

Anticipated Food Scarcity Scale (AFSS)

The Anticipated Food Scarcity Scale (AFSS), originally developed by Folwarczny et al. [5, 6], was used to measure individuals’ perceived level of anticipated scarcity regarding future food resources. The AFSS consists of 8 items rated on a 7-point Likert scale (1 = strongly disagree to 7 = strongly agree), with total scores ranging from 8 to 56. Higher scores indicate a stronger perception that food scarcity will occur in the future. Example items include “Food resources will become increasingly scarce” and “There may not be enough food for everyone [5, 6].

The Turkish version of the AFSS was adapted and psychometrically tested by our research group in a separate validation study (unpublished study). The translation process followed standard forward–backward translation procedures and expert panel review to ensure semantic and conceptual equivalence.

Content validity indices demonstrated strong agreement (I-CVI = 0.87–1.00; S-CVI/Ave = 0.98). Construct validity was supported by exploratory (KMO = 0.934; Bartlett’s χ²(28) = 2012.71, p < 0.001) and confirmatory factor analyses, confirming a unidimensional structure that explained 80.14% of the variance (factor loadings = 0.844–0.926; CFI = 0.999; TLI = 0.998; SRMR = 0.037).

The scale demonstrated excellent reliability in this sample (Cronbach’s α = 0.97; ICC = 0.838–0.912, p < 0.001), indicating strong internal consistency and temporal stability. These results confirm that the Turkish AFSS is a valid and reliable tool for assessing perceptions of anticipated food scarcity among university students and the general population. Scores were treated as continuous variables in the regression analyses.

Adult Food Security Scale (AFSSM)

In the current analysis, the Adult Food Security Survey Module (AFSSM) was used to assess participants’ and households’ access to adequate and balanced food. The AFSSM is the validated Turkish adaptation of the U.S. Adult Food Security Survey Module originally developed by the United States Department of Agriculture [18] and recently validated for Turkish university students [19].

The scale consists of eight items and evaluates individuals’ food security status over the past 12 months. Items are rated using mixed 3-, 4-, and 5-point Likert response options (e.g., “often true,” “sometimes true,” “never true”). The total possible score ranges from 0 to 48, with lower scores (0–10) indicating food security and higher scores (11–48) indicating food insecurity, as determined by ROC curve analysis in the Turkish validation study [19]. Factor analyses in the Turkish validation identified three dimensions—Inadequate Nutrition, Economic Concern, and Hunger—which together explained 77.4% of the total variance. The scale demonstrated good psychometric properties (Cronbach’s α = 0.769 for the total scale,subdimensions ranged from 0.742 to 0.761). Test–retest reliability was confirmed with significant correlations (p < 0.001) after a two-week interval [19].

This validated version was selected for the current research to ensure cultural and linguistic relevance. The scoring procedure and interpretation were standardized following Açar et al. [19] to facilitate comparability with global datasets [18–22].

Dietary intake assessment and calculation of the Nutrition-Based EAT index (NB-EAT)

To evaluate participants’ adherence to sustainable and balanced eating habits, a semi-quantitative Food Frequency Questionnaire (FFQ) and the Nutrition-Based EAT Index (NB-EAT) were utilized.

Dietary intake assessment

Daily dietary intake data were collected using a validated 52-item semi-quantitative Food Frequency Questionnaire [23, 24] adapted for Turkish adults. The FFQ assessed the frequency, portion size, and single-serving amounts of commonly consumed food groups such as dairy products, fruits, vegetables, grains, meats, legumes, oils, and beverages. Reported frequencies and portion sizes were multiplied by appropriate coefficients and converted into average daily intakes using the BEBIS (Nutrition Information System) software, which is based on Turkish food composition tables.

The original validation study conducted by Güneş et al. [23] demonstrated that the Turkish FFQ has an adequate level of validity for estimating habitual food and nutrient intake among adults. The energy-adjusted correlation coefficients ranged between r = 0.20–0.61, attenuation coefficients ranged between 0.34–0.66, and the quartile agreement with repeated 24-h dietary recalls exceeded 95%. These findings confirm that the FFQ is a suitable instrument for evaluating food and nutrient consumption in the Turkish population.

Calculation of the NB-EAT

The NB-EAT is a nutrient-based adherence index developed based on the EAT–Lancet Commission Report on Healthy Diets from Sustainable Food Systems [4]. The index evaluates the extent to which individuals’ daily nutrient intakes align with sustainable and balanced dietary targets.

The index includes 12 key nutrients, each coded as 1 (target met) or 0 (target not met), producing a total score ranging from 0 to 12. Higher scores indicate greater adherence to sustainable dietary principles.

The nutrient targets used were as follows: α-linolenic acid (≥ 2.5 g), carbohydrates (≥ 317.3 g), cholesterol (≤ 125.2 mg), dietary fiber (≥ 42.9 g), unsaturated fat (≥ 75.9 g), protein (≥ 90.1 g), saturated fat (≤ 22.7 g), total fat (≤ 105.6 g), calcium (≥ 717.8 mg), magnesium (≥ 732.5 mg), potassium (≥ 4100.7 mg), and added sugar (≤ 31.0 g).

Based on the total score, adherence levels were categorized as low (0–4), moderate (5–8), and high (9–12). Since NB-EAT is grounded in standardized nutrient assessment rather than psychometric testing, it ensures reproducibility and comparability across populations [1, 3, 4], in a study conducted with 5,285 adults in Türkiye, reported that food security was one of the key determinants of sustainable dietary motivations and, together with ecological footprint awareness, explained approximately 27% of the variance in food choices. This finding is consistent with our study, where food security emerged as the strongest predictor [15], on the other hand, showed in a study with university students that nutrition literacy directly and positively influenced sustainable dietary behaviors [16]; in that study, it was shown that individuals with higher nutrition literacy were more willing to pay for healthier foods even under food insecurity conditions. Conducted with 3,321 participants in South Korea, the study revealed that economic status was the strongest determinant of purchasing behaviors, but nutrition literacy facilitated healthier choices. This suggests that nutrition literacy may play a supportive role in transforming structural opportunities into sustainable dietary behaviors [13], it was reported that food insecurity and limited nutrition literacy among university students had negative impacts on diet quality and sustainable dietary behaviors. This finding indicates that our sample also represents one of the vulnerable groups in the higher education context.

Statistical analysis

All analyses were performed using IBM SPSS Statistics 22.0. During data entry, five male respondents and 58 participants with incomplete questionnaires were excluded, resulting in a final analytic sample of 1,200 female participants. Each questionnaire was manually verified; missing or inconsistent responses were clarified through direct follow-up, and no missing data were included in the final dataset.

The normality of continuous variables was assessed using the Kolmogorov–Smirnov test, skewness–kurtosis values, and Q–Q plots. Descriptive statistics (mean, SD, frequency, percentage) were used to summarize participant characteristics (Tables 1, 2 and 3).

Table 1.

Sociodemographic characteristics of the participants (n = 1200)

Variable	Mean ± SD
Age (years)	20.79 ± 1.97
Height (cm)	164.21 ± 6.64
Weight (kg)	59.73 ± 10.16
Number of main meals	2.43 ± 0.52
Number of snacks	1.30 ± 0.98
Body Mass Index (kg/m2)	22.14 ± 3.33
AFSSM_total	23.14 ± 7.66
AFSSM_inadequate nutrition	9.75 ± 4.72
AFSSM_economic anxiety	10.19 ± 3.45
AFSSM_hunger	3.20 ± 1.16
AFSS_total	43.34 ± 13.31
EINLA_total	29.68 ± 2.94
EINLA _Atotal (general information)	8.35 ± 1.48
EINLA _Btotal (reading comprehension)	5.55 ± 0.85
EINLA _Ctotal (food groups)	9.71 ± 0.63
EINLA _Dtotal (portion information)	1.63 ± 0.88
EINLA _Etotal (label reading and numerical literacy)	4.44 ± 0.88
NB_EAT_total	4.19 ± 1.04

AFSSM Adult Food Security Survey Module, AFSS Anticipated Food Scarcity Scale, EINLA Eating Information and Nutrition Literacy Assessment, NB_EAT Nutrition-Based Eating Adherence Index. Higher scores in AFSSM and AFSS indicate greater insecurity and scarcity perception; higher EINLA and NB_EAT reflect higher literacy and sustainable adherence

Table 2.

Distribution of sociodemographic and contextual variables

Variable	Category	n	%
Field of Education	Health	376	31.28
	Non-health	824	68.72
Marital Status	Married	16	1.29
	Single	1184	98.71
Nutrition Knowledge Source	School courses	172	14.32
	Health personnel	541	45.14
	Books	128	10.72
	Family/Friends	149	12.43
	Dietitian	406	33.84
Family Income (TRY)	< 7500	1177	98.14
	≥ 7500	23	1.86
Individual Income (TRY)	< 1000	794	66.24
	≥ 1000	406	33.76

TRY Turkish Lira. Percentages calculated based on valid responses. Nutrition knowledge refers to the individual’s main source of daily nutrition information

Table 3.

Distribution of Participants by Scale Categories

Category	Frequency	Percentage (%)
Food security (AFSSM)
Food security risk	1181	98.42
No food security risk	19	1.62
Nutrition literacy (EINLA)
Insufficient	1	0.1
Limited	51	4.32
Sufficient	1148	95.72
Nutrition-Based EAT Index (NB_EAT)
Low	778	64.84
Medium	419	34.93
High	3	0.3

Higher NB_EAT scores indicate stronger adherence to sustainable dietary patterns

To identify predictors of sustainable and balanced dietary preferences, a hierarchical regression model was applied.

• Model 1 included covariates (age, BMI, family income, nutrition expenditure, and educational field).

• Model 2 added food security (AFSSM) and perceived scarcity (AFSS).

• Model 3 introduced nutrition literacy (EINLA_total).

Multicollinearity was examined using Tolerance and VIF values (all < 2.0). Regression assumptions (linearity, homoscedasticity, and residual normality) were verified via residual and Q–Q plots. The explanatory power of each model was evaluated using R² and ΔR² values, with standardized β, 95% CI, and p-values reported (Table 4). AFSSM subdimensions were moderately correlated (r = 0.41–0.58, p < 0.001), indicating no collinearity concerns.

Table 4.

Hierarchical regression analysis results – factors predicting sustainable nutrition preferences

Model	Variable	ΔR2	R2	F	p	Significant variable(s)
Model 1	Control variables (BMI, age, nutrition expenditure, family income level)	—	0.006	1.512	0.183	Health-based education, p = 0.040
Model 2	+ Psychosocial factors (AFSS, AFSSM)	+ 0.007	0.013	2.162	0.035	AFSSM total score is significant p < 0.05
Model 2A	+ Psychosocial factors AFSSM sub-dimensions (malnutrition, economic anxiety, hunger)	+ 0.003	0.016	2.163	0.022	Malnutrition subdomain, p = 0.002
Model 3	+ Cognitive factors EINLA_total (Nutrition literacy)	+ 0.002	0.018	2.119	0.021	EINLA total is not significant p > 0.05
Model 3A	+ Cognitive factors EINLA_subdomains	+ 0,023	0,023	1.984	0.016	Nutrition literacy subdomains are not significant p > 0.05

ΔR²: Additional variance explained relative to the previous model

R²: Total explanatory power of the model

Model 2A: A submodel in which the subdimensions of food security (AFSSM) are added separately

BMI: Above 25 kg/m² and below 20 kg/m²; 20–25 kg/m²

Age: 20 years and below; 21 years and above

Education: Non-health field; Health field

Nutrition expenditure: Below 1000 TRY; other

Family income: Below 7500 TRY; above 7500 TRY

AFSS Perceived scarcity scale, AFSSM: Food security scale, EINLA Nutrition Literacy Scale

ΔR² = additional variance explained relative to the previous model; R² = model explanatory power. All models satisfied regression assumptions (normality, homoscedasticity, VIF < 2). Significance level set at p < 0.05.Higher AFSSM scores indicate greater food insecurity. Higher AFSS scores indicate stronger perception of future scarcity. Higher EINLA and NB_EAT scores reflect higher nutrition literacy and greater adherence to sustainable dietary patterns, respectively

The moderation analysis tested the effect of nutrition literacy on the association between food security and sustainable dietary preferences. Variables were mean-centered before creating interaction terms. The moderation model was significant (R² = 0.012, F = 4.696, p = 0.003). Food security predicted sustainable preferences (β = 0.097, 95% CI [0.045–0.149], p = 0.001), while the interaction term approached significance (β = 0.051, p = 0.082). A simple slopes analysis at low (–1 SD), medium (0), and high (+ 1 SD) levels of nutrition literacy confirmed a positive trend. The Johnson–Neyman test showed no additional significant regions beyond ± 1 SD.

To ensure robustness, a sensitivity analysis was conducted by re-running the models with additional sociodemographic covariates (urban/rural origin, smoking, physical activity), which did not change the results. The significance level was set at p < 0.05 for all analyses.

Results

Descriptive

Table 1 presents descriptive statistics for the 1,200 female university students included in the study. Participants had a mean age of 20.79 ± 1.97 years and an average BMI of 22.14 ± 3.33 kg/m². The mean NB_EAT score, reflecting sustainable and balanced nutrition preferences, was 4.19 ± 1.04. Mean food security (AFSSM_total) and anticipated food scarcity (AFSS_total) scores were 23.14 ± 7.66 and 43.34 ± 13.31, respectively. The nutrition literacy (EINLA_total) mean score was 29.68 ± 2.94.

More than two-thirds of participants (68.7%) were studying in non-health fields, 98.7% were single, and 91.8% reported no chronic disease. Regarding economic characteristics, most participants (98.1%) had household incomes below 7500 TRY, corresponding to the 2023 minimum wage in Türkiye. These findings indicate a relatively homogeneous, low-income, young adult sample predominantly outside health-related education (Table 2).

Table 3 shows categorical distributions of food security, nutrition literacy, and sustainable dietary adherence. The majority of participants (98.4%) were at some level of food insecurity risk. Nutrition literacy was adequate in 95.7% and borderline or insufficient in 4.3%. In terms of sustainable eating, 64.8% demonstrated low adherence to sustainable dietary behaviors.

Hierarchical regression

Determinants of sustainable and balanced dietary preferences were examined using a hierarchical regression model based on the NB_EAT score. The analyses were conducted in three stages. The control variables included in the first model (field of education, income level, age, body mass index) did not contribute significantly to the model (R² = 0.006, p = 0.183), but NB_EAT scores were found to be statistically higher in individuals with a health-related field of education (p = 0.040). Among the structural factors included in the model at the second level, the food security variable made the model significant (R² = 0.013, p = 0.035). In Model 2 A, which focused on sub-dimensions, the perception of malnutrition emerged as a significant predictor in a positive direction for sustainable choices (β = 0.106, p = 0.004). The nutrition literacy variable added to the model at the third level did not contribute significantly to the overall model (p = 0.147). These findings indicate that the strongest predictor of sustainable and balanced dietary preferences is the perceived food insecurity related to undernutrition, and that nutrition literacy alone is not a significant predictor (Table 4).

Moderator effect

Perception of scarcity

In the regression analysis evaluating whether nutritional literacy (EINLA_total) moderates the relationship between perception of scarcity (AFSS_total) and sustainable dietary choice (NB_EAT), the overall fit of the model was found to be quite weak (R² = 0.003; p = 0.279). These findings indicate that the model is not statistically significant and that the relationship between the variables is very low. The relationship curves between food insecurity perception and NB_EAT, plotted according to Z_EINLA levels, are parallel and horizontal, indicating that the interaction does not differ significantly and that there is no meaningful trend in the relationship. In conclusion, nutrition literacy does not significantly moderate the relationship between scarcity perception and sustainable choices (Fig. 1).

Fig. 1.

Moderating effect of nutrition literacy on the relationship between perceived food scarcity and sustainable dietary adherence (NB_EAT). The slopes for low, medium, and high levels of nutrition literacy are almost parallel, indicating no significant moderation effect (p = 0.279). X-axis: Perceived Scarcity (Z-score; − 1 = Low, 0 = Medium, 1 = High). Y-axis: Predicted sustainable dietary adherence (NB_EAT). Higher AFSS scores indicate stronger perception of future scarcity, while higher EINLA and NB_EAT scores reflect greater nutrition literacy and stronger adherence to sustainable dietary patterns

Food security

To examine whether nutrition literacy regulates the relationship between food security and sustainable nutrition preferences, a multiple linear regression analysis including an interaction term was applied. The overall model was found to be significant, F(3, 1196) = 4.696, p = 0.003, R² = 0.012.

Food security (β = 0.097, p = 0.001) significantly predicted sustainable nutrition preferences.

The interaction between food security and nutrition literacy revealed a finding at the threshold of significance (β = 0.051, p = 0.082), indicating a potential moderating effect (Table 5).

Table 5.

Moderating effect of nutrition literacy on the relationship between food security (AFSSM_total) and Sustainable Choice (NB_EAT)

Variable	B	β	p	Interpretation
Z(AFSSM_total)	0.100	0.097	0.001	Significant positive effect
Z(EINLA_total)	0.041	0.040	0.170	Not significant
Z(AFSSM × EINLA)	0.046	0.051	0.082	Marginal (trend-level)
Model fit:	F(3,1196) = 4.696, p = 0.003, R2 = 0.012

Interaction term (×) represents the moderating role of nutrition literacy. The model explains 1.2% of the variance in sustainable dietary choices. Given the marginal p-value (0.082), results should be interpreted with caution

Moderation analyzes were conducted using SPSS to assess the moderating role of nutrition literacy (Z-scored) on the relationship between food security and the NB_EAT score. Upon finding the interaction term significant, a simple slopes analysis was performed. In this analysis, the effect of food security on NB_EAT was tested separately at low (–1 SD), medium (average), and high (+ 1 SD) levels of nutrition literacy. According to the results, as the level of food security increases, the NB_EAT score significantly rises; the strength of this relationship becomes more pronounced as the level of nutrition literacy increases. In the line graph created to visualize the direction and magnitude of the interaction effect, the predicted NB_EAT values are presented for each level of nutrition literacy (Z = –1, 0, + 1) (Fig. 2).

Fig. 2.

Moderating effect of nutrition literacy on the relationship between food security and sustainable dietary adherence (NB_EAT). Steeper slopes are observed at higher literacy levels, indicating a marginally significant interaction (p = 0.082). X-axis: Food Security (Z-score; − 1 = Low, 0 = Medium, 1 = High). Y-axis: Predicted sustainable dietary adherence (NB_EAT). Higher AFSSM scores indicate greater food insecurity, while higher EINLA and NB_EAT scores reflect greater nutrition literacy and stronger adherence to sustainable dietary patterns

The ımpact of food security on ınsufficient and sufficient nutrition literacy groups

In this subgroup analysis, separate regression models were established for the Inadequate and Adequate Nutrition Literacy groups. No statistically significant relationship was found between food security and sustainable eating preference (NB_EAT) among individuals with inadequate nutrition literacy (B = –0.010, p = 0.463). This situation suggests that the sustainable preferences of individuals in this group may be independent of their level of food security. In contrast, in individuals with sufficient levels of nutrition literacy, food security emerged as a significant and positive predictor of sustainable dietary choices (B = 0.013, p = 0.001). This result suggests that nutrition literacy is not merely a level of knowledge, but is also related to individuals’ ability to respond more sensitively and consciously to environmental and structural factors (Table 6).

Table 6.

The ımpact of food security on sustainable dietary choice (NB_EAT) – subgroup analysis by nutrition literacy

Nutrition literacy	B	p	R2	Interpretation
Insufficient + Limited (n = 52)	–0.010	0.463	0.011	Non-significant
Sufficient (n = 1148)	+ 0.013	0.001	0.009	Significant positive effect

Only the sufficient group showed a significant association between food security and sustainable dietary adherence

Discussion

This study examined the predictive role of food security, perceived anticipated food scarcity, and nutrition literacy in sustainable dietary choices among female university students. The findings revealed that food security was the most significant determinant of sustainable and balanced dietary choices assessed within the scope of the study. In contrast, no significant relationship was found between perceived anticipated food scarcity and sustainable choices; nutrition literacy, while not a significant predictor on its own, played a moderating role by strengthening the relationship between food security and sustainable dietary choices.

The results obtained are largely consistent with the literature emphasizing the critical importance of food security in shaping sustainable dietary behaviors. Irgat et al. [15]. Mortaş et al. [16]. Similarly, another study conducted with young adults in Ankara reported that women had higher nutrition literacy scores, which were positively associated with behaviors such as seasonal food consumption, avoidance of food waste, and reduced meat consumption. In parallel, Selçuk et al. [14] also found that food insecurity was negatively associated with multiple components of sustainable and healthy eating behaviors—including adherence to a balanced diet, preference for regional and organic foods, seasonal consumption, and food waste avoidance—thereby supporting the argument that food insecurity substantially undermines sustainable dietary practices in adults. This evidence further situates our findings within the Turkish context, demonstrating a consistent pattern between limited food access and reduced engagement in sustainability-oriented nutrition behaviors. However, in our study, nutrition literacy was not a significant predictor on its own, but gained importance only in its interaction with food security. Therefore, while our findings parallel some studies in the literature (e.g., the strong role of food security as a determinant), they diverge in certain respects. In particular, the insufficiency of nutrition literacy alone suggests that structural factors play a more dominant role than individual cognitive capacities in shaping sustainable dietary behaviors. This indicates that improving individual knowledge and awareness alone may not be sufficient for sustainable diets; economic, social, and structural conditions must also be strongly considered.

The finding that nutrition literacy strengthened the relationship between food security and sustainable choices is noteworthy. This result is consistent with the findings reported by Nam et al. [13]. Conceptually, this finding also aligns with the “intersection of food literacy and sustainability” approach proposed by McManus et al. [25]. The authors systematically reviewed 114 studies conducted between 2013 and 2022 and demonstrated that the knowledge, skills, and attitude/behavior dimensions of sustainable food literacy could be strengthened through education. The moderator effect identified in our study supports the practical implications of this conceptual framework. In addition, the research conducted by O. MS [25] also directly addressed the role of nutrition literacy in sustainable diets, emphasizing that individual cognitive factors should be evaluated together with structural conditions [25].

Recent research has addressed the relationships between sustainable nutrition, food security, and nutrition literacy across different samples. Selçuk et al. [14] revealed that food insecurity among adults was negatively associated with behaviors such as balanced nutrition, preference for organic/regional products, seasonal consumption, and waste reduction. Nam et al. [13] also reported that even under food insecurity conditions, individuals with high nutrition literacy were inclined to pay for healthier foods. Similarly, Mengi Çelik et al. [26] demonstrated in adolescents that higher nutrition literacy was positively associated with adherence to the Mediterranean diet and sustainable environmental attitudes, which in turn reduced ecological footprint. In another study conducted at the higher education level, Hammad et al. [27] reported that graduate students and postdoctoral researchers experiencing food insecurity had poorer diet quality, characterized by lower consumption of vegetables, fruits, and whole grains, and higher consumption of refined products and sugary drinks. Taken together, these studies indicate that both structural factors (food security, economic capacity) and individual cognitive factors (nutrition literacy, environmental attitudes) shape sustainable dietary behaviors in a multidimensional way. Our study supports this literature by demonstrating that food security is the strongest determinant, while nutrition literacy plays a critical moderating role that strengthens this relationship.

The lack of a significant effect of perceived anticipated food scarcity on sustainable choices in our study is noteworthy. This suggests that the perception of scarcity reflects individuals’ cognitive evaluations of the future but may not directly guide daily dietary behaviors. The literature on this topic is limited; for example, Folwarczny et al. [5, 6] reported that scarcity perception creates more feelings of anxiety and insecurity in individuals, but its behavioral consequences may depend on contextual factors. Our findings support this evaluation by showing that perceived scarcity alone did not lead to concrete dietary changes in young adults.

The fact that our study was conducted only among female university students aligns with the literature emphasizing the importance of food insecurity and nutrition literacy in higher education. Indeed, in a comprehensive review by Haileslassie et al. [11]

When evaluated in terms of gender differences, Helvacı et al. [12] found that sustainable food literacy was higher particularly among women and individuals with a positive attitude toward healthy eating in young adults, while adherence to the Mediterranean diet was only weakly related to this literacy. Similarly, in their study conducted with adolescents, Mengi Çelik et al. [26] reported that girls had higher nutrition literacy and sustainable environmental attitudes than boys, while boys had a higher ecological footprint. These studies demonstrate that demographic characteristics (especially gender) are important variables to consider in shaping sustainable dietary behaviors and literacy levels.

The mechanisms underlying these associations should also be considered. Food insecurity may hinder sustainable eating behaviors among students through structural barriers such as higher costs of sustainable options, time limitations, and restricted cooking facilities. In contrast, nutrition literacy can help individuals overcome these constraints by improving their ability to read labels, plan meals, and make cost-effective sustainable choices. As all data were self-reported, potential common method and social desirability biases should be acknowledged, and the cross-sectional design prevents causal inference. University-level interventions that combine “literacy and structural support”—for example, affordable sustainable menus, curriculum-integrated literacy modules, and campus incentives—may provide more effective outcomes.

Strenght and limitations

Among the strengths of this research are the large sample size and the holistic evaluation of structural and cognitive factors such as food security, famine perception, and nutritional literacy within the same model. The fact that the perception of scarcity, a particularly current concept, has been examined in such a large sample provides a unique contribution to the literature. The study revealed that nutrition literacy alone is not a determinant of sustainable choices, but it plays a significant role in interaction with food security, indicating that stronger outcomes are achieved when individuals’ knowledge level is combined with their capacity to adapt to structural constraints. The finding in subgroup analyzes that individuals with sufficient nutritional literacy gave more sensitive and informed responses to food security conditions is an important finding supporting the need to consider individual and environmental factors together. In these aspects, the study not only contributes theoretically but also guides the more effective planning of public health policies that support sustainable dietary behaviors. Additionally, the economic dependency of the participants on their families ensured a relatively homogeneous socioeconomic structure, allowing for a clearer evaluation of the relationships among nutrition literacy, food security, and sustainable dietary choices. Furthermore, the exclusive inclusion of female participants can be considered a strength, as women generally exhibit higher nutrition literacy and sustainability awareness, allowing for a more focused analysis of these interrelated variables.

Due to the cross-sectional design, relationships between variables are shown, but causality has not been established. Since the scarcity perception scale measures future-oriented anxiety, it is expected to show a limited direct relationship with behavioral outcomes. The uneven distribution in nutrition literacy groups may have reduced statistical power, particularly due to the small size of the “inadequate” group. For this reason, the nutrition literacy variable was analyzed in two categories (insufficient and sufficient) to ensure adequate sample size and statistical validity. Moreover, the fact that the participant group consisted only of female university students limits the generalizability of the findings. However, the fact that the sample consisted exclusively of female university students also represents a limitation, as it restricts the generalizability of the findings to broader and gender-diverse population. In addition, since the participants were economically dependent on their families, the findings may not fully reflect the behaviors of economically independent adults.

Conclusion

This study examined structural (food security, perceived scarcity) and cognitive (nutrition literacy) factors influencing sustainable food choices in female university students using a hierarchical regression model. Analyzes have shown that food security is a significant determinant of sustainable dietary choices. Although the term “interaction” remains at the limit of significance, it suggests that nutrition literacy can play a regulatory role. Among the control variables, only individuals with health-related education were found to have a higher sustainable nutrition trend. Among structural factors, the perception of food security, particularly regarding nutritional adequacy, is a significant predictor, while the dimensions of hunger and economic concern were not found to be effective. This situation indicates that the perception of “food security” is prominent in sustainable choices. Subgroup analyzes revealed that food security only had a significant impact on sustainable preferences among individuals with adequate nutrition literacy. This result highlights that not only food access but also levels of information and awareness shape behavior. However, the perception of an impending food shortage did not have a significant impact on sustainable choices. This finding can be explained by the fact that the perception of famine reflects future anxieties and uncertainties, and therefore does not directly guide daily behavior. It is also thought that this perception could lead to defensive consumption patterns (e.g., choosing cheap, durable products), which could contradict sustainability principles. Generally, it is observed that sustainable eating behaviors are influenced by both environmental conditions (food security) and individual cognitive capacities (nutrition literacy). Therefore, policies supporting sustainable nutrition should not only focus on increasing food access; they should also be supported by education and awareness programs aimed at improving nutritional literacy.The findings of this research indicate that sustainable food choices are influenced not only by structural factors but also by individual cognitive abilities. Specifically, the relationship between food security and sustainable choices becomes meaningful in individuals with adequate levels of nutrition literacy, necessitating a multi-dimensional approach to policy and intervention recommendations. Strengthening nutrition literacy: Public health programs can play a supportive role in promoting sustainable dietary behaviors, especially among groups at high risk of food insecurity. Integration of food security policies: Efforts to increase food access may be more effective in terms of behavioral change when integrated with education and awareness programs that enhance cognitive capacity. Considering psychological constructs: It should be taken into account that uncertainty-based constructs like scarcity perception can lead individuals to short-term and defensive consumption patterns; intervention and communication strategies should account for these risks.

Authors’ contributions

Conceptualization: O.C., H.Y.; Writing – original draft: O.C., H.Y.; Writing – review and editing: O.C., H.Y.; Experimental work: O.C., H.Y.; Data collection: O.C. H.Y.; Visualization: O.C.; Supervision: H.Y., *Authors read and approved the final manuscript*

Funding

This study was conducted by the author(s) using their own resources, without financial support from any public, private, or commercial institution.

Data availability

The corresponding author can be contacted via e-mail upon reasonable and justified request.

Declarations

Ethics approval and consent to participate

Ethical approval for this research was obtained from the Gazi University Ethics Committee (February 9, 2023-E.5851006). The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. Informed consent was obtained from all participants prior to their participation in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.