The Interplay of Food Insecurity, Resilience, Stress Mindset, and Mental Distress: Insights From a Cross‐Sectional Study

ABSTRACT

Background and Aims

In the United States, food insecurity (FI) is a serious health issue potentially affecting brain function. While neuroimaging suggests that diet quality influences brain functions, gaps remain regarding its impact on resilience, stress mindset, and mental distress, particularly across age and gender. This cross‐sectional study investigated these relationships using data from 1099 participants, of whom 26.19% were females and 70.39% were males, with the majority (70%) being under 30 years.

Methods

A multi‐scale questionnaire assessing FI, resilience, stress mindset, and mental distress was distributed via social media. ANOVA and Ordinary Least Squares (OLS) regression were used to analyze the data in Python.

Results

FI was linked to reduced resilience and increased mental distress (p < 0.05), but did not produce an effect on stress mindset. Age, gender, education, and physical activity influenced neurobehaviors (p < 0.01), with physical activity showing the greatest improvement in resilience. Women exhibited stronger correlations between FI and neurobehaviors than men.

Conclusion

Encouraging physical activity and targeted mental health interventions can enhance resilience and reduce distress, particularly in women. Community‐based programs addressing gender and age disparities may be key to improving mental well‐being.

Abbreviations

ANOVA

analysis of variance

BRS

Brief Resilience Scale

CFA

confirmatory factor analysis

CFI

comparative fit index

DHA

docosahexaenoic acid

EPA

eicosapentaenoic acid

FI

food insecurity

HPA

hypothalamic‐pituitary‐adrenal

K‐10

Kessler Psychological Distress Scale10

PFC

prefrontal cortex

QR code

quick response code

SEM

structural equation modeling

SMS

Stress Mindset Scale

SR

stress response

TLI

Tucker‐Lewis index

1. Introduction

One of the most prevalent health problems in the US is food insecurity (FI), which affects over 44 million individuals annually, 30% of whom are children [1]. Additionally, during a 12‐month cycle, it is estimated that about 14 million households are food insecure at one point [2]. FI describes a provisional or permanent lack of sufficient food, which is essential for healthy living. A wide spectrum of causes ranging from political volatility to geographical location, and poverty could lead to food insecurity in communities. Food‐insecure individuals suffer from a wide range of food‐related issues, including malnutrition, nutritional‐related body dysfunctions, and nutritional deficiencies [3]. A lack of nutritious foods is often the source of this dietary disturbance, which increases the chance of enduring persistent ailments [4]. The connection between FI and mental health has only recently surfaced. Several dietary factors contribute to brain chemistry, neuroplasticity, and neurogenesis, which explains this association [5, 6]. Water‐ and fat‐soluble vitamins, choline, the omega‐3 fatty acids eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), and essential amino acids, are potentially low in food‐insecure individuals. Although FI is mostly identified in certain populations, it may cluster heavily among young adults due to the transition into financial independence. In addition, individuals on a budget are more likely to sacrifice the quality of their food to maximize their purchasing power. Their food choices hover over a myriad of ultra‐processed foods. The stress response (SR) could be impacted by the high fat and sugar contents of these foods [7]. Psychological stress is characteristically associated with FI and is a trigger for mental health decline. With the lingering of SR, physiological adaptations occur at the cellular and epigenetic levels. Psychological stress is characteristically associated with FI and is a trigger for mental health decline. With the lingering of SR, physiological adaptations occur at the cellular and molecular levels. The activation of the hypothalamic‐pituitary‐adrenal axis stimulates the release of glucocorticoids from the adrenal cortex, which bind to brain steroid receptors to regulate gene expression. These receptors, abundant in emotion‐regulating regions like the prefrontal cortex (PFC) and limbic system, influence the behavioral stress response. Individuals exposed to extreme stress often demonstrate higher resilience, reducing the likelihood of anxiety and depression, though most studies focus on older adults rather than young adults [8, 9]. Resilience is the capacity to mitigate the adverse effects of stress on psychological and physical well‐being, enabling individuals to confront adversity with a positive mindset and emotional control. It stems from active neurological and neuroendocrine adaptations to maintain stable functioning under stress [10]. A high‐quality diet has been associated with psychological resilience [11], but its influence on the stress mindset remains unclear, and the relationship between these traits and food insecurity is not well understood. Research on FI and cognitive functions has primarily examined different attributes such as attention, perception, memory, and speed of reaction among children and adults as reviewed by [12].

The rise of personalized approaches and precision medicine adds complexity to understanding the human brain and its associated neurobehaviors, including resilience, stress mindset, and mental distress. Although men and women share similar brain anatomy, differences in brain morphology result in indistinct behavioral responses [13]. These gender‐related morphological differences have been related to behavioral traits and susceptibility to mood disorders [14, 15]. For instance, women are more likely to be diagnosed or experience longer depression or anxiety episodes [14, 15, 16, 17]. Diet quality has a differential effect on the brains of men and women, leading to differences in mental health and brain function [18]. Preliminary evidence suggests gender‐based differences in neurobehaviors [19, 20]; however, disparity in context of food insecurity has not been explored. Brain maturity is another critical factor to consider, as the human brain continues developing into the mid to late 20 s, with the prefrontal cortex (PFC) being the last part to mature. PFC is responsible for emotional regulation and rationalization of thoughts, which accounts for the observed behavioral differences between young and mature adults. FI may impair brain development in young adults leading to negative mental and behavioral outcomes such as anxiety, depression, and suicidal ideation [12, 21]. However, the impact of FI on neurobehaviors in young versus mature adults remains poorly understood.

Considering the various unknowns across multiple factors, such as the differential effects of diet, gender, age, and food insecurity on neurobehaviors, there is a significant gap in understanding how these variables interact to influence outcomes such as resilience, stress mindset, and mental distress. Therefore, the objective of this study is to investigate the interactions between diet, gender, age, and food insecurity, and their collective impact on neurobehaviors such as resilience, stress mindset, and mental distress. This study expands the understanding of food insecurity by linking it to neurobehavioral traits. It also addresses critical gaps in the literature by focusing on the unique vulnerabilities of young adults (due to ongoing brain development) and gender (due to brain morphological differences).

2. Methods

2.1. Participants

This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects were approved by the Binghamton University Internal Review Board. Adults 18 years or older were invited to participate in the study. A built‐in informed consent form was provided to all potential participants at the beginning of the survey. Participation was voluntary and participants were able to accept or decline to take part in this study by selecting the appropriate answer choice. Those who declined to participate exited the survey. Participants were given the chance to skip any question they did not feel comfortable answering. All participants consented to the study before they accessed the survey. All data collection occurred electronically.

2.2. Study Design

This cross‐sectional study utilized an anonymous survey using Google Forms. The survey was distributed using flyers with a QR code and various social media platforms such as Instagram, Snapchat, Reddit, Facebook, and GroupMe. The survey link also targeted college students as most of the young adults are enrolled in colleges. The inclusion criterion was adults of 18 years or older. Data collection occurred between August 2022 and March 2023. Demographic questions collected information on age, gender, level of education, and engagement in physical activity for at least 20 min a day. Four validated scales were used, which included topics on Food Insecurity, Resilience, Stress Mindset, and Mental distress. The Food and Agriculture Organization of the United Nations‐Food Insecurity Experience Scale [22] consists of eight questions that require a dichotomous response of “Yes” or “No.” It was used to measure household food security by evaluating a wide range of behaviors that describe food insecurity. The Brief Resilience (BRS) scale [23] assessed the level of resiliency to adversity and the perceived ability to bounce back. BRS encompasses six items with items 1, 3, and 5 positively phrased whereas items 2, 4, and 6 have a negative connotation. The BRS is scored by reverse coding items the negative items (2, 4, and 6). Participants were asked to rate their level of agreement with statements using a 5‐point Likert scale ranging from 1 to 5, with 1 being “Strongly Disagree” and 5 being “Strongly Agree”. The Stress Mindset Scale (SMS) [24], is an eight‐item tool that depicts beliefs about the nature of stress as enhancing or debilitating. Like BSR, SMS uses a 5‐point Likert scale ranging from 1 to 5, with 1 being “Strongly Disagree” and 5 being “Strongly Agree” and contains four negative items which are scored by reverse scoring. Higher scores on the SMM represent the mindset that stress is enhancing. The Kessler Psychological Distress Scale (K10) [25] measures psychological distress using ten questions related to emotional state. K10 uses a 5‐point Likert scale as well ranging from “None of the time” to “All of the time”. High scores indicate high levels of psychological distress. Gender was partitioned based on responses stated Men or Women. All other responses were not included in the analysis. Age partitioning was based on the suggested age of brain maturity completion [26]. Young adults were considered 18‐29 years old, and mature adults ≥ 30 years of age.

2.3. Statistical Design and Analysis

2.3.1. Rationale of the Statistical Design

The study used a multi‐analyses approach to investigate the individual effect of FI on each neurobehaviors and the effects of demographics on FI and neurobehaviors by partitioning the data into men and women.

The first step used Factor Analysis to reduce the dimensionality of the data since the research survey was a multi‐scale. A correlational analysis followed and generated a heat map displaying the strength and direction of the relationships between FI and neurobehaviors.

Statistical analysis was performed through confirmatory factor analysis (CFA) to analyze the measurement model, Analysis of Variance (ANOVA) for group comparisons and to identify the association between FI and neurobehaviors with demographics, and Ordinary Least Squares (OLS) regression to test associations. The last step involved a base model approach to classify the findings from the regression analysis based on gender while exploring the age group differences. All two‐sided tests were conducted at the α  = 0.05 level of significance. Results are reported with p‐values following standard conventions (p < 0.001, p < 0.01, p < 0.05), alongside effect sizes. For the ANOVA test, the effect size known as “eta squared” (η²) is used, which quantifies the proportion of variation in the dependent variable that is explained by the independent variable. A value below 0.01 is negligible, 0.01–0.06 is small, 0.06–0.14 is medium, and 0.14 or higher is large, indicating the increasing influence of the independent variable on the dependent variable [27]. ANOVA and regression were pre‐specified analyses; gender‐stratified regression was exploratory. Analyses were performed in Python version 3.8 with STATA models and Pandas libraries.

2.4. Factor Analysis

Minimum sample size with 95% confidence and a 5% margin of error was set at 385 responses. However, to boost the study power, data collection went beyond the minimum sample size. The survey included four different scales Food Insecurity, Resilience, Stress Mindset, and Kessler Psychological Distress Scale (K10), which made the data multidimensional. Structural Equation Modeling (SEM) is a powerful statistical technique used to examine and estimate causal relationships between variables by integrating elements of multiple regression and factor analysis. SEM involves two primary stages [1]: CFA to validate the measurement model, which ensures that observed indicators accurately measure underlying latent factors, and [2] structural model evaluation, which tests the hypothesized relationships between those latent factors, including causal pathways and indirect effects contributing to theoretical insights [28]. To simplify the complexity of the data set, CFA was used to manage its multidimensional nature, which is a widely used method in social and behavioral sciences [29], including food and nutrition studies [3, 30]. In this study, these latent factors represent the four main constructs of interest identified earlier (Food Insecurity, Resilience, Stress Mindset, and Kessler Psychological Distress Scale). The hypothesized CFA model's validity was confirmed through goodness‐of‐fit indices, enhancing the robustness of latent construct measurement while addressing potential biases from the online survey methodology Similar items were grouped into distinct factors via CFA, as illustrated in Figure 1. As the foundation of SEM, CFA validates the measurement model using goodness‐of‐fit indices, including the chi‐square statistic, the comparative fit index (CFI), the Tucker‐Lewis index (TLI), and the root‐mean‐square error of approximation (RMSEA) [31]. This validation ensures the constructs are measured reliably, enabling researchers to confidently proceed to the structural model for testing theoretical relationships.

Figure 1.

CFA diagram showing two latent variables (circles) measured through six observed indicators (rectangles Y1–Y6) with measurement errors. Latent variables represent unobserved constructs measured indirectly through these indicators.

If the indicators are {Y ₁, Y ₂,…, Y _n }, and the common factors are {F ₁, F ₂,…, F _m }, the overall SEM model with “m” factors can be represented by Equation 1.

$${\mathrm{Y}}_{\mathrm{i}}=\mu +{\lambda }_{\text{ik}}{\mathrm{F}}_{\mathrm{k}}+{\in }_{\mathrm{i}},$$ (1)

where μ represents the intercept or mean of the observed variable Y _i , λ _ik is the loading of the i ^th variable on the k ^th common factor, F _k is the score on the k ^th common factor latent, and ϵ _i is the i ^th specific factor error term [29]. The SEM model illustration is represented in Figure 2. The SEM equations for the Resilience latent variable (F₁) that have six observed variables (RS₁‐RS₆) can be written as:

$${\text{RS}}_1=\mathrm{\mu }+{\mathrm{\lambda }}_{11}{F}_1+{\mathrm{ϵ}}_1$$

$${\text{RS}}_2=\mathrm{\mu }+{\mathrm{\lambda }}_{12}{F}_1+{\mathrm{ϵ}}_2$$

$${\text{RS}}_3=\mathrm{\mu }+{\mathrm{\lambda }}_{13}{F}_1+{\mathrm{ϵ}}_3$$

$${\text{RS}}_4=\mathrm{\mu }+{\mathrm{\lambda }}_{14}{F}_1+{\mathrm{ϵ}}_4$$

$${\text{RS}}_5=\mathrm{\mu }+{\mathrm{\lambda }}_{15}{F}_1+{\mathrm{ϵ}}_5$$

$${\text{RS}}_6=\mathrm{\mu }+{\mathrm{\lambda }}_{16}{F}_1+{\mathrm{ϵ}}_6$$

Figure 2.

The latent variables (circles) include Resilience, SM (Stress Mindset), FI (food insecurity), and MD (mental distress). The observed variables (rectangles) measure these constructs using multiple indicators: Resilience (RS1–RS6), SM (SM1–SM8), FI (FI1–FI8), and MD (MD1 to MD10). Solid arrows represent direct effects, with factor loadings and p values indicating significance, while dashed arrows represent indirect or mediating effects.

Factor loadings (λ) are directly represented by the coefficients in Figure 2 (e.g., λ ₁₁ = 0.754 for RS₁). A factor score was calculated for each record by summing up observed intakes of factor items weighted by factor loading, and each score was normalized.

2.5. Analyses of Variance

To investigate whether significant differences exist between the means of the main four factors of Food Insecurity, Resilience, Stress mindset, and Mental Distress among different subgroups of gender, age, current student, education level, and physical activity for the sample (n = 1099), a one‐way ANOVA was conducted. ANOVA is a statistical method used to compare the variation of means between groups [30]. A significant p < 0.05 in ANOVA results indicates that there is at least one pair of subgroups in which the mean difference is statistically significant, and this socio‐demographic characteristic affects the corresponding factor. The assumption of equal variance in the ANOVA test between subgroups was examined using Levene's test. In case the assumption of equal variance was not met, denoted by (^a) in the ANOVA table, we conducted Welch's ANOVA, which is effective in dealing with unequal variances, and reported the results accordingly [32]. Welch's ANOVA was used to reduce bias from unequal variances, ensuring more reliable group comparisons despite imbalances in representation and online participation. Gender, age, and student variables were coded as nominal such as gender (Male, Female), age (age < 30, age ≥ 30), and Student (Yes, No), while education level, and physical activity, variables were coded as ordinal such as education level (High school or less, Bachelor, Master, and PhD), and physical activity (0–2 times/week, 3–5 times/week, 6–7 times/week). The age breakdown (age < 30 as young adults and age ≥ 30 as mature adults) was based on the brain maturity level as suggested by Somerville [26]. The N/A or Not Sure options were excluded from the analysis.

2.6. Regression Analysis

To assess the exact impact of socio‐demographic characteristics on Food Insecurity, Resilience, Stress Mindset, and Mental Distress, an Ordinary Least Squares (OLS) regression analysis was conducted using a base model for each subgroup (denoted as ^b in Table 3). OLS regression enables us to determine the statistical significance of estimated coefficients, providing insights into the direction and strength of variable relationships. The importance of the OLS model is evaluated using the p‐value associated with the F‐statistic [33]. The base model provides a consistent benchmark for comparison, enabling us to assess the contribution of individual characteristics to the observed factor across subgroups. Based on this base model, the specific effects of socio‐demographic variables on well‐being factors were examined. The regression coefficient for each subgroup shows its contribution relative to the base group.

Table 3.

Regression analyses of the association of socio‐demographic characteristics with all factors.

Variable	Food insecurity coefficient (95% CI)	Resilience section coefficient (95% CI)	Stress mindset coefficient (95% CI)	Mental distress coefficient (95% CI)
Gender
Male (b)
Female	0.0012	−0.259***	−0.197***	0.174***
	(−0.008, 0.010)	(−0.348, −0.170)	(−0.279, 0.116)	(0.102, 0.247)
Age
Age < 30 (b)
Age ≥ 30	−0.018***	0.182***	−0.169***	−0.270***
	(−0.029, −0.008)	(0.076, 0.289)	(−0.265, −0.074)	(−0.354, −0.187)
Currently Student
No (b)
Yes	0.006	−0.067	0.158***	0.105**
	(−0.002, 0.015)	(−0.166, 0.014)	(0.079, 0.238)	(0.034, 0.177)
Education Level
HS (b)
Bach	0.0006	0.162***	−0.009	−0.093**
	(−0.008, 0.009)	(0.078, 0.248)	(−0.085, 0.067)	(−0.162, −0.026)
Graduate	−0.0113	0.065	−0.231***	−0.090
	(−0.026, 0.004)	(−0.086, 0.218)	(−0.369, −0.094)	(−0.213, 0.032)
PA
0‐2/wk (b)
3‐5/wk	−0.010**	0.135**	0.194***	−0.127**
	(−0.019, −0.002)	(0.049, 0.222)	(0.114, 0.275)	(−0.200, −0.056)
6‐7/wk	−0.011**	0.319***	0.258***	−0.175***
	(−0.023, −0.001)	(0.215, 0.434)	(0.155, 0.363)	(−0.268, −0.082)

3. Results

3.1. Demographics

A total of 1099 participants completed the survey, of which 26.19% were females and 70.39% were males. The majority of the respondents (70%) were less than 30 years old. Of the respondents, 71.97% were students, with 53.95% in high school or below, 38.22% holding a bachelor's degree, and 7.82% holding a master's or PhD. Additionally, 48.86% reported moderate levels of physical activity (Table 1).

Table 1.

Descriptive statistics of socio‐demographic characteristics (n = 1099).

Covariate	Frequency	Percent (%)
Gender
Male	298	26.19
Female	801	70.39
Age
Age < 30	899	79.00
Age ≥ 30	239	21.00
Currently student
No	319	28.03
Yes	819	71.97
Education Level
High School or Less	614	53.95
Bachelor's	435	38.22
Master's & PhD	89	7.82
Physical activity (20 min of physical activity)
0–2 times/week	373	32.78
3–5 times/week	556	48.86
6–7 times/week	209	18.37

3.2. SEM Model

The SEM diagram comprises several fundamental components that clarify the structural relationships between variables. Figure 2 shows the overall SEM model with latent and observed variables. These observed variables provide measurable data points that help quantify the underlying constructs represented by the latent variables.

Factor loadings represent how strongly each observed variable relates to its corresponding latent variable (e.g., 0.754 for RS₂, p < 0.00).

According to the SEM model fit, the RMSEA was 0.05, and its 90% confidence interval (CI) interval (0.047, 0.067. Steiger [34] considers RMSEA values < 0.10 as an acceptable model fit. Both CFI (0.86) and NNFI (0.85) approach 0.9, which is an acceptable value for a model. The ratio of the chi‐squared to degrees of freedom (χ2/df) = 1.3 < 2, indicates a very good fit. However, the (χ2) values were also highly significant, which is common in large data sets (Burr, 1988). The results of the survey showed a reasonably well‐fitting model where the fit indices (RMSEA, CFI, TLI, SRMR) are all within acceptable ranges; therefore, supporting the SEM model used in the study.

Figure 2 shows that the absolute standardized factor loadings were significant at different p values (0.05,0.01,0.001) and varied within the following ranges: 0.45–0.78 for the food insecurity factor, 0.63–0.76 for the resilience factor, 0.44–0.75 for the stress mindset factor, and 0.54 to 0.81 for the mental distress factor.

3.3. Factors Correlation

The correlation coefficients range from −1 to 1, with a coefficient of 1 indicating a perfect positive correlation, a coefficient of −1 indicating a perfect negative correlation, and a coefficient of 0 indicating no correlation between the variables. The significant correlation for p < 0.05 between the four factors ranges from 0.15 to 0.69 as shown in Figure 3.

Figure 3.

The correlation matrix displays relationships among food insecurity, resilience, stress mindset, and mental distress. A red‐to‐blue gradient indicates correlation strength and direction, with red for positive and blue for negative correlations. Diagonal cells represent the correlation of each variable with itself and show perfect correlations (1.00), while off‐diagonal cells present correlation coefficients. Asterisks denote significance levels: p < 0.05, p < 0.01, p < 0.001.

The correlation analysis reveals significant interesting relationships between different factors. First, food insecurity exhibits moderately significant positive correlations with mental distress (r = 0.37) and a small negative correlation (r = −0.15) with resilience, while no significant association was depicted with the stress mindset. Resilience produced a significant positive correlation with stress mindset (r = 0.41) and a significant negative correlation with mental distress (r = −0.69). The stress mindset produced a significant negative correlation (r = −0.32) with mental distress.

3.4. ANOVA Test

A one‐way ANOVA test with α = 0.05 was conducted to determine if there are significant differences in the means of Food Insecurity, Resilience, Stress mindset, and Mental Distress among different subgroups such as Gender, Age, current student, education level, and physical activity (n = 1099) (Table 2).

Table 2.

ANOVA analysis of socio‐demographic variables.

Variable	Food insecurity				Resilience				Stress mindset				Mental distress
	Mean	p value	p‐ꭓ2	Effect size (η²)	Mean	p value	p‐ꭓ2	Effect size (η²)	Mean	p value	p‐ꭓ2	Effect size (η²)	Mean	p value	p‐ꭓ2	Effect size (η²)
Gender
Male	−0.001	0.905	0.488	0.0001	0.15	p  <  0.001	0.564	0.029	0.15	p  <  0.001	0.871	0.021	−0.013	p  <  0.001	0.462	0.022
Female	−0.003				−0.04				‐0.04				0.031
Age
Age < 30	0.002	p  <  0.001	p  <  0.001 a	0.012	−0.03	p  <  0.001	0.133	0.012	0.030	p  <  0.001	0.769	0.011	0.040	p  <  0.001	0.814	0.036
Age ≥ 30	−0.014				0.16				−0.13				−0.220
Student
No	−0.004	0.147	0.06	0.002	0.044	0.15	0.133	0.002	−0.11	p  <  0.001	0.419	0.014	−0.07	p  <  0.001	0.074	0.008
Yes	0.002				−0.019				0.04				0.03
Edu Level
HS	0.0001	0.392	0.198	0.002	−0.07	p  <  0.001	0.484	0.013	0.021	p  <  0.001	0.858	0.010	0.044	p  <  0.001	0.639	0.010
Bach	0.001				0.094				0.012				−0.049
Graduate	−0.010				−0.004				−0.209				−0.046
PA
0–2/wk	0.0060	0.020	0.641	0.006	−0.129	p  <  0.001	0.020 a	0.027	−0.142	p  <  0.001	0.348	0.027	0.049	p  <  0.001	0.549	0.017
3–5/wk	−0.003				0.008				0.051				−0.033
6–7/wk	−0.004				0.201				0.115				−0.080

Abbreviations: HS, high school; PA, physical activity.

^a Welch's ANOVA Bold values: significant at 0.05 level.

According to ANOVA results, age, and physical activity have a significant effect on food insecurity. For instance, the mean of response to food insecurity questions between age groups was 0.002 for age < 30 years and −0.014 for ≥ 30 years old group indicating a significant difference between subgroups (p < 0.05, η² = 0.012). The resilience factor is significantly impacted by all factors, except being a student. However, the stress mindset is impacted by all factors. Interestingly, all factors, gender, age, being a student, educational level, and physical activity affect mental distress.

3.5. Regression Analysis

To precisely study the effect of socio‐demographic characteristics subgroups on each factor, a regression analysis model was generated. All the socio‐demographic characteristics were used as control variables in the regression modeling while scores of the four factors (output of SEM model) were used as the dependent variables. Based on the regression results (Table 3), females had a significantly lower resilience and stress mindset by 0.259 and 0.197, respectively (p < 0.001), and an increase in mental distress by 0.174 compared to males. On the other hand, those ≥ 30 years of age had significantly decreased risk of food insecurity, lower stress mindset, and mental distress by 0.0186, 0.169, and 0.270, respectively with an increment in resilience by 0.182 compared to those who are less than 30 years of age (p < 0.001). Compared to nonstudents, being a student increases stress mindset by 0.158 and mental distress by 0.105 (p < 0.001). Education level produced significance between bachelor and graduate degrees in regard to resilience, stress mindset, and mental distress. Furthermore, the most interesting model is the one generated by PA at least three times a week. Physical activity (moderately active and very active) was significantly associated with lower food insecurity (−0.010, −0.011), higher resilience (0.135, 0.319), a more positive stress mindset (0.194, 0.258), and lower mental distress (−0.127, −0.175), respectively, when compared to non‐active individuals. PA was the only variable that showed positive attributes across the board, with resilience and to a lesser extent, showing improvement with a higher frequency of exercise. All significance we set at (p < 0.01) (Table 3).

For a deeper understanding of the impact of gender across all other independent variables, a regression interaction model was used, and the data set was stratified by gender. This approach examines the interplay between gender and specific conditions, providing insights into the effects of the factors studied. For instance, the interaction between females and the ≥ 30‐year‐old group describes that these mature females are less likely to be food insecure, with resilience and better mental well‐being, but with a more debilitating stress mindset by −0.014, 0.251, −0.134, and −0.308, respectively. Students females are less likely to be resilient, with a positive stress mindset, but more at risk of mental distress by −0.135, 0.133, and 0.145, respectively.

As with the regression model, the same trend for women and education was noted, with higher education likely to be associated with a more debilitating stress mindset by −0.205 compared to high school degrees. PA among females is strongly associated with resilience, with active individuals being less likely to be food insecure when compared to inactive females. The same trend was seen for resilience with a higher frequency of PA (Table 4).

Table 4.

Regression analysis results stratified by gender: Females subgroup.

Variables	Food insecurity coefficient (95% CI)	Resilience section coefficient (95% CI)	Stress mindset coefficient (95% CI)	Mental distress coefficient (95% CI)
Age
Age < 30b
Age ≥ 30	−0.014**	0.251***	−0.134**	−0.308***
	(−0.027, −0.003)	(0.132, 0.371)	(−0.244, −0.025)	(−0.404, −0.213)
Currently student
Nob
Yes	0.0054	−0.135*	0.133**	0.145***
	(−0.005, 0.015)	(−0.237, −0.033)	(0.041, 0.226)	(0.064, 0.227)
Education level
High Schoolb
Bach	−0.0008	0.199***	0.008	−0.110**
	(−0.010, 0.009)	(0.101, 0.297)	(−0.080, 0.098)	(−0.189, −0.032)
Graduate	−0.0067	0.071	−0.205**	−0.109
	(−0.023, 0.009)	(−0.079, 0.246)	(−0.357, −0.053)	(−0.089, 0.346)
PA
0–2/wkb
3–5/wk	−0.013**	0.143**	0.138**	−0.145***
	(−0.023, −0.003)	(0.042, 0.245)	(0.042, 0.245)	(−0.227, −0.063)
6–7/wk	−0.015**	0.301**	0.151**	−0.147**
	(−0.029, −0.002)	(0.164, 0.438)	(0.142, 0.295)	(−0.258, −0.038)

Note: Bold values: significant at (*p < 0.05; **p < 0.01; ***p < 0.001) level, ^b: Base model.

Abbreviations: HS, high school; PA, physical activity.

3.5.1. Males

The regression interaction model was repeated for males and its impact on the neurobehaviors of interest. This model produced less significant interactions when compared to women. Interestingly, none of the sociodemographic factors in males were associated with resilience. Mature males ( ≥ 30 years of age) in the cohort are less likely to be food insecure, with a more debilitating stress mindset (−0.313), and with a lower risk for mental distress (−0.1820) when compared to their younger counterparts. Being a male student and physically active was associated with an enhanced stress mindset by 0.201 and 0.317, respectively, compared to not being a student and being inactive. Additionally, males with a graduate degree or aged 30 years or older were found to have a more debilitating stress mindset (Table 5).

Table 5.

Regression analysis results stratified by gender: Males subgroup.

Variables	Food insecurity coefficient (95% CI)	Resilience section coefficient (95% CI)	Stress mindset coefficient (95% CI)	Mental distress coefficient (95% CI)
Age
Age < 30b
Age ≥ 30	−0.029**	0.0005	−0.313**	−0.182***
	(−0.051, −0.008)	(−0.215, 0.215)	(−0.510, −0.117)	(−0.352, −0.012)
Currently student:
Nob
Yes	0.013	0.054	0.201**	0.006
	(−0.005,0.031)	(−0.009, 0.062)	(0.037, 0.365)	(−0.082, 0.215)
Education level
High schoolb
Bachelor	0.003	0.0892	−0.027	−0.032
	(−0.013, 0.020)	(−0.052, 0.220)	(−0.530, 0.009)	(−0.034, 0.046)
Graduate	−0.024	−0.066	−0.451**	−0.054
	(−0.044, 0.030)	(−0.059, 0.010)	(−0.771, −0.131)	(−0.077, 0.030)
PA
0–2/wkb
3–5/wk	−0.0041	−0.0149	0.317**	0.022
	(−0.024, 0.015)	(−0.059, 0.003)	(0.142, 0.493)	(−0.140, 0.185)
6–7/wk	−0.0027	0.191	0.428**	−0.058
	(−0.025, 0.020)	(−0.280, 0.354)	(0.226, 0.631)	(−0.060, 0.009)

Note: Bold values: significant at (*p < 0.05; **p < 0.01; ***p < 0.001) level, ^b: Base model.

Abbreviations: HS, high school; PA, physical activity.

4. Discussion

The purpose of the study was to fill a gap by assessing the impact of food insecurity on resilience, stress mindset, and mental distress. Another gap addressed was the effect of FI on neurobehaviors regarding gender and age groups. The hypothesis was that FI has a positive effect on resilience, a negative effect on mental distress, and a more debilitating impact on stress mindset. The study revealed several interesting findings, although some were unexpected. First, food insecurity may lower resilience and promote mental distress, without having an impact on the stress mindset, which was highly unexpected (Figure 3). We hypothesized that these attributes are neuro‐related, and they allegedly should move in the same direction. Second, gender, age groups, education, and physical activity modulate levels of resilience, stress mindset, and mental distress. Third, physical activity produced a strong positive impact on all neurobehaviors, but resilience saw the most significant improvement with a higher frequency of exercise. Finally, many of these associations were detected among females and to a lesser extent in men, as predicted; however, young adults showed a mix of neurobehaviors, which was not expected.

4.1. The Impact of Food Security on Neurobehaviors

The correlational results, as illustrated by the heat map (Figure 3) were expected for mental distress, but not for resilience and the stress mindset. The literature describes that resilience develops from adversity, so the anticipated results were that FI is positively associated with resilience, as food insecurity typically comes with hardship. However, looking into the neurobiology of resilience through work performed on laboratory animals [10], resilience develops from neural and molecular adaptations when the organism is exposed to stress. The stress response induces the activation of the hypothalamic‐pituitary‐adrenal (HPA) axis, generating a cascade of extensive hormonal, physiological, and neurochemical intermediates [35]. An effective activation of the HPA axis is believed to support the development of resilience, while a dysfunctional HPA axis prevents it. The literature also describes that a nutrient‐dense diet, such as a Mediterranean diet style, promotes psychological resilience, which proposes the role of diet quality in resilience development [36]. It is putative that several nutritional metabolites are involved in the molecular processes of resilience. Therefore, nutrient deficiency, as experienced with FI may explain our findings. Previous work on psychological resilience and nutrition looked into the effect of resilience on emotional eating [37], adherence to healthy eating [38], and snaking behaviors [39]. Therefore, this finding proposes a novel concept that is worth further exploration.

According to Crum [24], individuals have different assessments of a stressful situation; it could be viewed as debilitating or enhancing. Lazarus and Folkman (1984) proposed that an initial appraisal is necessary to determine the extent of the stress demands and a second appraisal to assess the coping ability to handle the stressor [40]. A stress‐is‐debilitating mindset views the situation as threatening when the demands offset the coping ability. On the other hand, the stress‐is‐enhancing mindset views stress as a motivator to achieve. The fact that food insecurity was not associated with the stress mindset, could be related to other factors such as personality, genetics, and environment among others. Additionally, the mind is dynamic and can be easily influenced by external factors, which may explain the lack of association with FI. The relationship between diet quality and mental health is more transparent than the other neurobehaviors. Diet quality has been linked to the modulation of mental health, therefore it is expected that FI may increase the risk of mental distress.

4.2. Relationship of FI and Neurobehaviors With Demographics

The ANOVA analysis shed light on the impact of the socio‐demographic factors on FI and neurobehaviors suggesting that gender has no impact on FI. Those ≥ 30 years old showed a significantly inverse relationship with FI. As expected, the highest physical activity levels were negatively associated with FI. The literature contains mixed results about the association between age and food insecurity. Some reports propose that food insecurity hovers around midlife [41], while others reported a higher association with younger age [42]. However, the discrepancy could be related to geographical, racial, and health factors. The relationship between FI and physical activity was expected. Those who experience FI are less likely to be physically active, which confirms other reports in the literature [43]. Interestingly, females produced a significant inverse relationship with resilience, which also supports findings from the literature [20]. Women are less likely to control their emotions due to their higher amygdala activity [44]. Living with negative emotions may explain the reason women are more likely to be diagnosed with depression than men [14].

Physical activity (PA) has a significant effect on resilience. PA modulates neurotransmitters and supports neuroplasticity [45], which is essential for building resilience. Additionally, those who are physically active tend to consume a high‐quality diet supporting the connection between diet quality and resilience [46]. The findings related to the stress mindset were interesting. The stress‐is‐debilitating mindset was associated with being a female, ≥ 30 years old, and with a graduate degree. This combination of factors suggests that potential temporal burnout may be associated with this phenomenon. Although being female is not necessarily associated with a time‐based trait, women are more likely to ruminate negative thoughts and juggle several responsibilities that eventually lead to burnout [47, 48].

4.3. Predictive Models of FI and Neurobehaviors

The regression analysis was used to support the ANOVA results and predict the influence of demographics on FI and neurobehaviors. Those ≥ 30 years old and those who are moderate to highly active are less likely to be food insecure. The regression analysis revealed that women have lower resilience, a more debilitating stress mindset, and higher mental distress when compared to men, which aligns with several reports in the literature. The stress mindset is typically linked to psychological well‐being [49] and life satisfaction [24]. It is known that women are more likely to be influenced by a stress‐is‐debilitating mindset than men [50]. Because women to tend have a more negative appraisal of situations, it makes it harder to control negative emotions [51]. Another fascinating outcome is that moderate‐high physical activity is associated with higher resilience, a stress‐is‐enhancing mindset, and lower mental distress, while there is a negative association with food insecurity, which confirms the importance of diet in building these positive attributes. Arida et al. [52] proposed that exercise builds brain reserves that empower the brain's neurochemicals and structures to withstand the negative effects of stress. Exercise induces the release of serotonin, dopamine, and endorphins which collectively improve mental wellbeing, and support a positive mindset [53]. The resilience factor is significantly impacted by all demographic factors, except being a student. Similarly, the stress mindset is impacted by all factors. Interestingly, gender, age, being a student, educational level, and physical activity affect mental distress.

It was expected that lower resilience would always coincide with higher mental distress. Our results confirmed that resilience and mental distress move in the opposite direction, even when the dataset was stratified by gender. It was also expected that stress mindset and resilience co‐occur; however, the results of this study show that this is not always the case (Figure 4). The instance when resilience co‐occurred with a debilitating stress mindset was for ≥ 30 years of age, which suggests a more factual outlook on life matters with age. The opposite scenario of having low resilience with a positive mindset was detected among female students, which suggests motivation and hope for the future.

Figure 4.

(A) The hypothesized relationship between FI and neurobehaviors. (B) The depicted relationship between FI and neurobehaviors. The asterisks represent the discrepancy.

4.4. Future Direction

Future research should focus on longitudinal studies to better understand the long‐term effects of diet, food insecurity, and their interaction with gender and age on neurobehaviors. Additionally, interventional trials are needed to evaluate the efficacy of nutritional strategies in mitigating the negative neurobehavioral impacts of food insecurity. Exploring these avenues will provide crucial insights into how to address the emerging mental health concerns associated with food poverty across different populations.

4.5. Study Strengths and Limitations

The study has several strengths and limitations. The large sample, filling a gap in the literature, and the robust analysis are the strengths of the study. The cross‐sectional nature of the study does not establish a cause‐and‐effect relationship, which is a limitation. Also, the accuracy of the responses is inherently not verified in an anonymous survey, which is another limitation. Moreover, bias in internet access may have excluded individuals who are less engaged with digital platforms, thereby restricting generalizability.

5. Conclusion

The study filled several gaps in the literature and provided a comprehension outlook on the relationship between FI and several neurobehaviors. The current study provided a novel concept that diet quality may moderate the development of resilience. Additionally, gender and age groups, based on brain maturity) may experience different levels of resilience, stress mindset and mental distress, which necessitates categorization of these variables for future research.

Author Contributions

Lina Begdache: conceptualization, investigation, methodology, project administration, supervision, writing – original draft, writing – review and editing. Amera Al‐Amery: data curation, formal analysis, software, writing – original draft. Katerina K. Nagorny: data curation, investigation, methodology. Ushima Chowdhury: data curation, investigation, methodology. Lexis R. Rosenberg: data curation, investigation, methodology. Zeynep Ertem: resources, software, validation, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Transparency Statement

The lead author Lina Begdache affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.