Association between ultra-processed food consumption and odds of depression, stress, and anxiety in adolescent girls: a cross-sectional study

Abstract

Background and aim

Mental disorders, including depression, anxiety, and stress, disproportionately affect adolescent girls, with dietary patterns emerging as modifiable risk factors. Ultra-processed food (UPF) consumption, prevalent among adolescents, may exacerbate these conditions. This study examines the association between UPF consumption and mental health outcomes in Iranian female adolescents.

Methods

In this cross-sectional study, 263 female adolescents aged 15–18 years from Tehran, Iran, were assessed. Dietary intake was measured using a validated 168-item food frequency questionnaire, with UPFs classified per the NOVA system and expressed as a percentage of total energy intake. Depression, anxiety, and stress were evaluated using the DASS-21 questionnaire. Multivariable logistic regression models estimated odds ratios (ORs) and 95% confidence intervals (CIs) across UPF consumption tertiles, adjusting for BMI and physical activity.

Results

Higher UPF consumption was significantly associated with increased odds of depression, stress, and anxiety. In fully adjusted models, participants in the highest UPF tertile had 3.69 times higher odds of depression (95% CI: 1.92–7.10, P trend < 0.001), 2.84 times higher odds of stress (95% CI: 1.49–5.41, P trend = 0.001), and 1.99 times higher odds of anxiety (95% CI: 1.10–3.66, P trend = 0.026) compared to the lowest tertile.

Conclusion

Greater UPF consumption is associated with elevated odds of depression, stress, and anxiety in adolescent girls, underscoring the need for dietary interventions to mitigate mental health burdens in this vulnerable population.

Introduction

Mental health disorders rank among the leading causes of disease burden in adolescents worldwide, affecting approximately 10–20% of this age group, with depression and anxiety being particularly prevalent [1]. In Iran specifically, studies have documented that approximately 30–35% of adolescents experience symptoms of depression and anxiety, with rates being notably higher among female adolescents compared to their male counterparts [2]. Notably, adolescent girls face substantially higher rates of these conditions compared to boys [3]. These disparities stem from complex biological and psychosocial factors, including hormonal changes during puberty, body image pressures, and social expectations [4]. When left untreated, these conditions can derail academic performance, damage social relationships, and increase risks for substance abuse and chronic psychiatric disorders in adulthood [5]. Traditional approaches to understanding adolescent mental health have emphasized genetic and psychosocial factors, but recent evidence points to dietary patterns as important modifiable risk factors [6]. Studies indicate that nutrient-dense, anti-inflammatory diets associate with lower depression and anxiety risk, whereas diets high in refined carbohydrates and low in essential nutrients correlate with worse mental health outcomes [7].

Ultra-processed foods (UPFs), as defined by the NOVA classification system, are industrially manufactured products composed of five or more ingredients, often incorporating substances such as additives, preservatives, emulsifiers, and artificial flavorings that are not typically found in home cooking [8]. Global consumption of UPFs has risen dramatically over recent decades, with adolescents representing the highest consumers across age groups [9]. Their engineered palatability can activate brain reward systems similar to addictive substances, potentially promoting compulsive eating and emotional dysregulation [10, 11].

A growing body of epidemiological evidence supports associations between UPF consumption and adverse mental health outcomes, primarily in adult populations [12]. A recent meta-analysis found that higher UPF consumption was associated with a 22% increased odds of depression (HR = 1.22, 95% CI: 1.16–1.28) among adults [13]. Dose-response analyses have revealed that ultra-processed food consumption exhibits a threshold relationship with depression odds, where even a shift from low to moderate intake (approximately 400 g/day) significantly increases the odds, but additional consumption beyond this level does not substantially elevate it further. This suggests that even moderate ultra-processed food consumption may be sufficient to adversely affect mental health outcomes [14].

However, studies specifically examining these associations in adolescent populations remain scarce, with even fewer focusing on adolescent girls despite their heightened vulnerability. The limited available evidence in adolescents suggests similar trends to those observed in adults, with preliminary data indicating potential sex-specific effects that warrant further investigation. The present study addresses critical gaps in the current literature by focusing specifically on adolescent girls, a population simultaneously experiencing heightened vulnerability to mental health disorders and demonstrating concerning patterns of UPF consumption. Given the substantial burden of depression, anxiety, and stress among adolescent girls and the modifiable nature of dietary exposures, identifying potential associations between UPF consumption and mental health outcomes could inform targeted preventive interventions with significant public health implications.

Methods and materials

Study design and participants

This cross-sectional study investigated the association between UPF consumption and the odds of depression, stress, and anxiety among 263 female adolescents aged 15–18 years in Tehran, Iran. Participants were recruited using a multistage stratified cluster sampling approach, ensuring representation across three socioeconomic strata (affluent, middle-class, and deprived) as designated by the Iranian Ministry of Education. High schools were randomly selected from each stratum, and students were subsequently sampled randomly from these schools. Exclusion criteria included chronic illnesses, specifically diabetes mellitus, cardiovascular diseases (coronary heart disease, heart failure), chronic kidney disease, chronic liver disease, cancer, and severe psychiatric disorders (schizophrenia, bipolar disorder), use of psychotropic medications, adherence to specific diets, smoking, or a prior diagnosis of depression or anxiety. The study was conducted in accordance with the principles outlined in the Declaration of Helsinki and received ethical approval from the Ethics Committee of the National Nutrition and Food Technology Research Institute at Shahid Beheshti University of Medical Sciences, Tehran, Iran (ethics approval number 054577). Informed written consent was obtained from all participants prior to their inclusion in the study.

Depression, stress, and anxiety assessment

Mental health outcomes were evaluated using the reliable and Persian-validated Depression, Anxiety, Stress Scale-21 (DASS-21) [15, 16], a shortened version of the DASS-42 developed by Lovibond and Lovibond [17]. This questionnaire has demonstrated strong reliability (Cronbach’s alpha > 0.80 for all subscales) and validity in Iranian populations [15, 16]. The Persian version of the DASS-21 has demonstrated good reliability (Cronbach’s alpha > 0.80 for all subscales) and validity in Iranian populations across similar age groups. The DASS-21 comprises three subscales (depression, anxiety, and stress), each with seven items scored on a 4-point Likert scale. Total scores for each subscale were calculated by summing relevant items, with higher scores indicating greater severity. Participants were categorized as “healthy” (normal range) or “unhealthy” (mild to severe symptoms) based on established cutoffs: depression (≤ 9 vs. ≥10), anxiety (≤ 7 vs. ≥8), and stress (≤ 14 vs. ≥15).

Dietary assessment

Dietary intake over the preceding year was assessed using a validated, semi-quantitative food frequency questionnaire (FFQ) tailored for the Iranian population, comprising 168 food items with standard serving sizes [18]. This FFQ has been validated in Iranian adults, showing acceptable reproducibility and relative validity for major nutrients (correlation coefficients ~ 0.40–0.70) and its reliability has been confirmed through comparisons with 24-hour dietary recalls in prior studies [18, 19]. Participants reported consumption frequency (daily, weekly, or monthly), and responses were converted to daily grams using a household measures manual [20]. Energy and nutrient intakes were calculated using the USDA food composition database within Nutritionist 4 software, supplemented by the Iranian Food Composition Table for local foods not included in the USDA database [21].

UPF classification and quantification

UPFs were identified and quantified based on the NOVA classification system, which categorizes foods according to the extent and purpose of industrial processing. Each FFQ item was independently classified into one of the four NOVA groups by two trained nutrition researchers. Discrepancies were resolved through discussion with a third senior researcher until consensus was achieved. UPFs (NOVA Group 4) include products such as carbonated beverages, packaged snacks, instant noodles, ready-to-eat meals, processed meats (e.g., sausages, hot dogs), sweetened breakfast cereals, and commercially produced breads and cakes, characterized by high levels of added sugars, saturated fats, sodium, and additives (e.g., emulsifiers, artificial sweeteners) [8]. Each FFQ item was classified into one of the four NOVA groups, with a focus on Group 4 for this study. The contribution of UPFs to total energy intake (% kcal) was calculated by summing the energy content of all UPF items and dividing by total daily energy intake. Participants were then categorized into tertiles of UPF consumption (% kcal) for analysis, with the lowest tertile serving as the reference group.

Anthropometric and lifestyle variables

Trained staff performed all anthropometric assessments. Body weight was measured to the nearest 0.1 kg using a calibrated digital scale (Seca 881, Germany), with participants wearing light clothing and no shoes. Standing height was determined to the nearest 0.5 cm using a portable, non-elastic measuring tape. Body mass index (BMI) was subsequently calculated by dividing weight (kg) by height squared (m²). Physical activity levels were evaluated through a validated self-administered questionnaire and reported as metabolic equivalent hours per day (MET.h/day) [22, 23].

Statistical analysis

Data were analyzed using SPSS version 21.0 (IBM Corp., Chicago, IL, USA), with a significance threshold of p < 0.05. Normality of continuous variables was assessed using Kolmogorov-Smirnov tests and histograms; non-normal variables were log-transformed. Participant characteristics and dietary intakes across UPF consumption tertiles were compared using one-way ANOVA for continuous variables and chi-square tests for categorical variables. Multiple logistic regression models estimated odds ratios (ORs) and 95% confidence intervals (CIs) for depression, anxiety, and stress across UPF tertiles, with the lowest tertile as the reference. Two models were constructed: Model 1 (crude) and Model 2 (adjusted for BMI and physical activity). Trends across tertiles were evaluated by treating tertile medians as a continuous variable in regression models. All analyses accounted for potential confounders identified through literature review and preliminary data exploration.

Results

Participant characteristics

The mean ± SD age and BMI of participants were 16.2 ± 0.96 years and 22.3 ± 4.1 kg/m², respectively. Across tertiles of ultra-processed food (UPF) consumption, which ranged from < 22.07% (T1), 22.07–30.28% (T2), to ≥ 30.28% (T3) of total energy intake, no significant differences were observed in anthropometric measurements, including weight, height, BMI, waist circumference, and waist-to-hip ratio (Table 1). However, depression scores were significantly higher among participants in the highest tertile of UPF consumption (12.8 ± 10.6) compared to those in the lowest tertile (8.9 ± 9.3) (P = 0.019). There were no significant differences in anxiety and stress scores across UPF consumption tertiles. Total energy intake was significantly higher among participants in the highest tertile of UPF consumption (2824 ± 845 kcal/day) compared to those in the lowest tertile (2445 ± 783 kcal/day) (P = 0.002). As expected, the proportion of energy derived from UPFs progressively increased across tertiles, from 17.3 ± 3.4% in the lowest tertile to 36.0 ± 4.9% in the highest tertile (P < 0.001).

Table 1.

Characteristics of participants (n = 263) according to the UPFs consumption tertiles (% of Kcal)

	Tertile of total UPFs (% of Kcal)
	T1	T2	T3	P value
Age (y)	16.3 ± 0.95	16.1 ± 0.91	16.2 ± 1.03	0.263
Weight, kg	59.8 ± 11.5	59.1 ± 12.1	58.1 ± 11.8	0.660
Height, cm	162.3 ± 5.5	161.9 ± 6.1	163.5 ± 5.8	0.180
Body mass index, kg/m2	22.7 ± 4.2	22.5 ± 4.1	21.7 ± 4.1	0.253
Waist circumference (cm)	72.8 ± 7.5	73 ± 8.5	71.2 ± 8	0.256
Waist to hip ratio	0.75 ± 0.06	0.75 ±‌ 0.05	0.74 ± 0.04	0.290
Depression score	8.9 ± 9.3	9.6 ± 8.3	12.8 ± 10.6	0.019
Anxiety score	8.8 ± 8.1	8.3 ± 6.9	9.7 ± 8.3	0.464
Stress score	14.2 ± 9.4	13 ± 9.1	14.4 ± 9.3	0.562
Body image, %				0.913
Normal	77	76	76
Impaired	13	14	14
Physical activity (MET/h)	36.5 ± 5.4	36.4 ± 6.4	35.6 ± 5.4	0.480
Calorie intake (Kcal/day)	2445 ± 783	2443 ± 807	2824 ± 845	0.002
UPF consumption (% of TEE)	17.3 ± 3.4	26.6 ± 2.7	36 ± 4.9	< 0.001

Values are means ± SDs for continuous variables and percentages for categorical variables

ANOVA for quantitative variables and χ² test for qualitative variables

Dietary intakes across UPF consumption tertiles

Table 2 presents dietary intakes according to tertiles of UPF consumption. A significant inverse relationship was observed between UPF consumption and intake of unprocessed or minimally processed foods (NOVA group 1), with participants in the highest UPF tertile consuming significantly less unprocessed foods (1404 ± 510 g/d) compared to those in the lowest tertile (1728 ± 628 g/d) (P < 0.001). As expected, consumption of ultra-processed foods (NOVA group 4) progressively increased across tertiles, from 293 ± 175 g/d in the lowest tertile to 663 ± 354 g/d in the highest tertile (P < 0.001). No significant differences were observed in the consumption of processed culinary ingredients (NOVA group 2) and processed foods (NOVA group 3) across UPF consumption tertiles.

Table 2.

Dietary intakes of participants according to UPFs consumption tertiles (% of Kcal)

	Tertiles
	T1	T2	T3	P trend
unprocessed or minimally processed foods (NOVA group 1) (g/d)	1728 ± 628	1419 ± 551	1404 ± 510	< 0.001
processed culinary ingredients (NOVA group 2) (g/d)	53 ± 51	45 ± 30	49 ± 26	0.411
processed foods (NOVA group 3)	254 ± 120	255 ± 124	256 ± 125	0.992
ultra-processed food (NOVA group 4)	293 ± 175	418 ± 173	663 ± 354	< 0.001

Data are presented as mean ± SD

ANOVA test

Association between UPF consumption and mental health outcomes

A strong positive association was observed between UPF consumption and odds of depression (Table 3). In the crude model (Model 1), participants in the highest tertile of UPF consumption had 3.79 times higher odds of depression (95% CI: 1.99–7.23) compared to those in the lowest tertile. This association remained robust after adjustment for physical activity and BMI (Model 2: OR = 3.69, 95% CI: 1.92–7.0.92.0, P trend < 0.001). Similarly, UPF consumption was positively associated with odds of stress. In the fully adjusted model (Model 2), participants in the highest tertile of UPF consumption had 2.84 times higher odds of stress (95% CI: 1.49–5.41, P trend = 0.001) compared to those in the lowest tertile. This association remained significant after controlling for potential confounders. A significant association was observed between UPF consumption and odds of anxiety. In the crude model, participants in the highest tertile of UPF consumption had 1.95 times higher odds of anxiety (95% CI: 1.07–3.63) compared to those in the lowest tertile. After adjustment for physical activity and BMI (Model 2), this association remained significant (OR = 1.99, 95% CI: 1.1–3.66, P trend = 0.026). These findings demonstrate a consistent pattern of higher odds of depression, stress, and anxiety with increasing consumption of ultra-processed foods among adolescent girls, with the strongest association observed for depression.

Table 3.

Odds and 95% confidence interval for occurrence of the mental disorders in each tertile of UPF

Depression	Tertiles		P trend
	T1 (n = 78)	T2 (n = 88)	T3 (n = 98)
	(< 22.07)	(22.07–30.28)	(30.28 ≤)
No. of cases	21	35	58	< 0.001
Model 1	ref	1.73 (0.89–3.34)	3.79 (1.99–7.23)	< 0.001
P value		0.104	< 0.001
Model 2	ref	1.7 (0.87–3.35)	3.69 (1.92–7.92)	< 0.001
P value		0.108	<0.001
Stress
No. of cases	21	35	51	0.005
Model 1	ref	1.75 (0.91–3.38)	2.89 (1.52–5.39)	0.001
P value		0.104	0.001
Model 2	ref	1.72 (0.89–3.34)	2.84 (1.49–5.41)	0.001
P value		0.106	0.001
Anxiety
No. of cases	32	42	58	0.068
Model 1	ref	1.24 (0.66–2.3)	1.95 (1.07–3.63)	0.024
P value		0.471	0.026
Model 2	ref	1.26 (0.68–2.3)	1.99 (1.1–3.66)	0.026
P value		0.470	0.027

Based on multiple logistic regression model

Model 1: crude

Model 2: additionally adjusted for physical activity, BMI

Both the total number of participants and the number of cases (defined as mild to severe symptoms according to DASS-21 cutoffs) are reported for each tertile. Analyses were conducted on the full sample of 263 participants with complete data

Discussion

This cross-sectional study investigated the association between UPF consumption and mental health outcomes among adolescent girls in Tehran, Iran. Our findings revealed significant positive associations between UPF consumption and odds of depression, stress, and anxiety, with the strongest relationship observed for depression. Specifically, participants in the highest tertile of UPF consumption had 3.69 times higher odds of depression, 2.84 times higher odds of stress, and 1.99 times higher odds of anxiety compared to those in the lowest tertile, after adjusting for potential confounders.

Our findings are consistent with the expanding body of literature that associates higher consumption of ultra-processed foods with negative mental health outcomes. A systematic review and meta-analysis encompassing 385,541 participants reported that elevated intake of UPFs was linked to greater odds of experiencing depressive and anxiety symptoms in cross-sectional analyses, as well as a heightened odds of developing depression in longitudinal studies [24]. In a similar vein, a large prospective cohort study of 183,474 participants from the UK Biobank indicated that individuals in the highest quartile of UPF consumption (Q4) had a 26% and 11% increased odds of depression and anxiety, respectively, compared to those in the lowest quartile (Q1) over a median follow-up period of 13.1 years [25].

Our study demonstrated a particularly strong association between UPF consumption and depression in adolescent girls, with those in the highest tertile of UPF consumption having 3.69 times higher odds of depression compared to those in the lowest tertile. This robust association is consistent with findings from other populations. A study in the Korean general population revealed that females in the highest quartile of UPF intake had a 1.51 times higher likelihood of having depression, after adjusting for confounders [26]. Moreover, a dose-response meta-analysis indicated that for every 10% increase in ultra-processed food consumption, an 11% higher odds of depression was observed [27]. The relationship between UPF and stress observed in our study (OR = 2.84, 95% CI: 1.49–5.41) complements existing research on working-class young adults, which found that workers with high stress levels had higher levels of ultra-processed food consumption [28]. This bidirectional relationship suggests that stress may lead to increased UPF consumption, while UPF consumption may exacerbate stress symptoms. Regarding anxiety, our study found a significant association (OR = 1.99, 95% CI: 1.1–3.66) between UPF consumption and odds of anxiety. This finding adds to the mixed evidence in the literature, where some meta-analyses have reported significant associations between UPF intake and higher prevalence of anxiety [24], while others have not found a significant association [27].

Several biological and behavioral pathways may underlie the associations observed between UPF intake and adverse mental health outcomes. First, UPFs are often characterized by poor nutritional quality, lacking essential micronutrients such as vitamins and minerals due to a minimal presence of whole food ingredients like fruits and vegetables. This micronutrient inadequacy may contribute to an elevated odds of depression through multiple physiological mechanisms [29]. Second, a high intake of UPFs has been associated with alterations in gut microbiota composition (gut dysbiosis), which may impair gut-brain axis communication and diminish the synthesis of key neurotransmitters, including serotonin [27, 30]. Additionally, UPF consumption may disrupt hypothalamic-pituitary-adrenal (HPA) axis function, influencing appetite-regulating hormones and neurotransmitter systems that play a role in mood regulation [14, 31, 32]. Third, gut dysbiosis linked to elevated UPF intake may trigger the release of pro-inflammatory cytokines, leading to higher circulating levels of high-sensitivity C-reactive protein [33, 34]. Inflammation is known to influence the onset, progression, and treatment outcomes of mental disorders, and peripheral inflammatory markers have been proposed as potential biomarkers for these conditions [35]. Fourth, various additives prevalent in UPFs may interfere with the synthesis and regulation of neurotransmitters involved in mood, including dopamine, norepinephrine, and serotonin [36]. Furthermore, emulsifiers commonly used as antimicrobial agents (e.g., carboxymethylcellulose and polysorbate-80) may alter the gut microbiome, promoting inflammatory processes [37, 38]. Fifth, exposure to chemicals used in food packaging materials, such as Bisphenol A, has been linked to dysregulation of stress-responsive and endocrine systems, which may contribute to symptoms of anxiety and depression [39]. Similarly, titanium dioxide nanoparticles have been associated with increased levels of inflammatory cytokines in both the bloodstream and brain tissue, as well as neuroinflammatory responses [40]. Finally, the high energy density and palatability of UPFs, compared with minimally processed foods, may promote excessive energy intake by altering gut-brain signaling pathways and disrupting flavor-nutrient feedback mechanisms [41]. This pathway may help explain the higher total energy consumption observed among participants in the upper tertile of UPF intake in our study.

The association between stress and UPF intake warrants particular consideration, especially among adolescent girls, who may be more susceptible to emotional eating behaviors. Psychological stressors can significantly impact dietary patterns, with chronic stress being linked to increased consumption of energy-dense foods, as well as higher intakes of saturated fats and sugars [42]. In some individuals, stress not only promotes greater food intake but also serves as a key contributor to unhealthy dietary habits, a pattern that appears particularly prevalent among young adults [42]. UPFs, which are typically highly palatable, may elicit physiological effects that help attenuate stress by stimulating the endogenous opioid (reward) system and dampening the activation of the HPA axis [43, 44]. These foods may also alleviate stress through mechanisms such as sensory pleasure, distraction, or psychological escape, in addition to possible nutritional or metabolic influences [43, 44]. Such processes may help explain the tendency for individuals to consume UPFs as a coping strategy in response to stressors that are perceived as uncontrollable.

Implications for public health and clinical practice

The findings of our study carry meaningful implications for both public health policy and clinical care, particularly regarding adolescent mental health. In light of the significant associations observed between UPF consumption and elevated odds of depression, stress, and anxiety, it is advisable to incorporate these mental health outcomes when evaluating the disease burden linked to UPF intake and when formulating strategies to safeguard population health. This consideration is increasingly pertinent given the marked rise in the production, availability, and consumption of ultra-processed foods worldwide over recent decades, a trend that is projected to continue. Growing support exists for the application of the precautionary principle in addressing UPF consumption, as well as its production and distribution, within newly established dietary guidelines developed by national and international health agencies [45]. Our results align with this perspective and further strengthen existing observational and experimental evidence underscoring the benefits of adopting healthier dietary patterns for the prevention and management of mental health disorders. Nutritional interventions in clinical settings may contribute to alleviating the burden of mental health conditions among adolescent girls. Our findings suggest that improving dietary behaviors could play a dual role: enhancing mental well-being and promoting healthier food choices, indicating a potentially reciprocal relationship between diet quality and mental health promotion.

Strengths and limitations

This study possesses several methodological strengths. Firstly, dietary intake was assessed using a validated FFQ specifically adapted for the Iranian population, thereby improving the reliability of dietary data. Secondly, we applied the NOVA classification system, a widely accepted framework, to systematically categorize foods based on their degree of processing. Thirdly, our analyses accounted for key confounding variables, including BMI and physical activity levels. Nevertheless, some limitations warrant consideration. Foremost, the cross-sectional design of this study represents a fundamental limitation that precludes any causal inference between UPF consumption and mental health outcomes. Cross-sectional studies, by their nature, capture exposure and outcome data at a single time point, making it impossible to establish temporal precedence. Consequently, we cannot ascertain whether elevated UPF consumption preceded the development of depression, stress, and anxiety symptoms, or whether pre-existing mental health conditions influenced dietary choices toward greater UPF intake. The potential for reverse causation is substantial and particularly relevant in the context of mental health and dietary behaviors. Individuals experiencing depressive symptoms, stress, or anxiety may engage in emotional eating or seek comfort foods, which are often ultra-processed and hyper-palatable, as a coping mechanism. This behavioral response could artificially inflate the observed associations between UPF consumption and mental health outcomes. Furthermore, bidirectional relationships are highly plausible: poor mental health may drive increased UPF consumption through mechanisms such as reduced motivation for meal preparation, altered appetite regulation, or stress-induced eating behaviors, while simultaneously, chronic UPF consumption may contribute to the onset or exacerbation of mental health symptoms through the biological pathways. Without longitudinal data with repeated measures of both dietary intake and mental health status, we cannot disentangle these complex, potentially reciprocal relationships. Therefore, our findings should be interpreted as associations rather than causal relationships, and they serve primarily to generate hypotheses for future prospective investigations. Secondly, despite adjusting for several confounders, the possibility of residual or unmeasured confounding cannot be entirely excluded given the observational design. Socioeconomic factors such as educational attainment, household income, parental education, and employment status could potentially influence both dietary habits and mental health outcomes and were not fully captured in our analysis. Thirdly, although the FFQ employed has been validated within the Iranian context, it was not explicitly developed to quantify energy intake derived from ultra-processed foods. Moreover, there remains a risk of misclassification of UPF intake; however, such misclassification would likely be non-differential, potentially biasing the results toward the null.

Fourthly, measurement concern relates to the reliability of self-reported dietary intake among adolescents, a population known to be particularly prone to under-reporting and over-reporting food consumption. Several factors may compromise the accuracy of dietary recall in this age group. Social desirability bias may lead adolescent girls to under-report consumption of foods perceived as unhealthy or associated with weight gain, including many ultra-processed foods such as sweets, fast food, and sugar-sweetened beverages. Conversely, some adolescents may over-report intake of foods considered healthy to present themselves more favorably. Cognitive factors also play a role, as adolescents may have difficulty accurately estimating portion sizes, recalling eating occasions outside structured meal times (such as snacking), or remembering foods consumed over extended recall periods. These challenges may be particularly pronounced for UPFs, which are often consumed in unstructured settings, in variable portion sizes, and as snacks rather than meals. Additionally, adolescents may have limited awareness of the processing level of foods they consume, potentially leading to misclassification when attempting to distinguish between minimally processed and ultra-processed versions of similar foods. While the FFQ method employed in our study has been validated and helps standardize portion size estimation, it cannot entirely eliminate these sources of measurement error. If present, such misclassification would likely be non-differential with respect to mental health outcomes, meaning errors in dietary reporting would be unrelated to depression, anxiety, or stress status. Non-differential misclassification typically attenuates observed associations toward the null, suggesting that our findings might underestimate the true strength of relationships between UPF consumption and mental health outcomes rather than overestimate them. Nevertheless, the possibility of differential misclassification cannot be entirely ruled out if, for example, participants with depression systematically reported their diets differently than those without depression. Future studies employing objective dietary assessment methods or biomarkers of UPF consumption would help address these limitations. Moreover, mental health outcomes were assessed using the self-reported DASS-21 questionnaire. While this tool has been validated, it may not offer the same diagnostic precision as structured clinical interviews conducted by mental health professionals, potentially leading to misclassification or underestimation of mental health symptoms.

Our sample of 263 participants, while sufficient to detect the strong associations reported here, is relatively modest compared to large-scale epidemiological studies examining diet-mental health relationships. This limited sample size may have constrained our statistical power to identify weaker associations or to conduct more detailed stratified analyses by potential effect modifiers such as socioeconomic status, family structure, or pubertal stage. More critically, our exclusive focus on adolescent girls from Tehran, Iran, substantially limits the generalizability of our findings across multiple dimensions. While this gender-specific approach was intentional and scientifically justified, given that adolescent girls face disproportionately higher burdens of depression and anxiety, exhibit distinct dietary vulnerabilities, and may experience sex-specific biological and psychosocial pathways linking diet to mental health it inherently restricts the external validity of our conclusions. Our findings cannot be extrapolated to adolescent boys, who demonstrate different prevalence patterns of mental health disorders, divergent dietary behaviors, and potentially distinct neurobiological responses to dietary exposures. Furthermore, the geographic and cultural specificity of our sample introduces additional constraints on generalizability. Dietary patterns, food availability, cultural attitudes toward food and mental health, and socioeconomic determinants of both diet quality and mental health care access vary considerably across regions and cultures. Iranian adolescents may have unique dietary traditions, different patterns of UPF availability and consumption, and culturally specific stressors that influence mental health outcomes in ways that differ from adolescents in Western countries or other middle-income nations. Therefore, replication studies in diverse populations, including male adolescents, different age groups, varied geographic and cultural settings, and populations with different socioeconomic characteristics are essential to establish whether the associations we observed represent generalizable phenomena or are specific to the particular demographic we studied. Until such replication is achieved, our findings should be considered most applicable to adolescent girls in urban Middle Eastern settings with similar sociodemographic characteristics to our sample.

Future research directions

Given the cross-sectional nature of the present study, establishing causality between UPF consumption and mental health outcomes remains a critical priority for future research. Longitudinal cohort studies with multiple waves of data collection are essential to determine whether UPF consumption precedes the onset of mental health symptoms or whether pre-existing mental health conditions influence dietary choices. Such studies should incorporate repeated assessments of both dietary intake and mental health status at multiple time points, ideally beginning before the onset of symptoms, to disentangle temporal relationships and reduce the risk of reverse causation bias. Additionally, where ethically feasible, randomized controlled dietary intervention trials that modify UPF intake while monitoring changes in mental health outcomes would provide the most robust evidence for causal relationships. These experimental designs could help clarify whether reducing UPF consumption leads to improvements in depression, anxiety, and stress symptoms among adolescents.

Further investigations are warranted to deepen our understanding of the association between UPF consumption and mental health outcomes, particularly in adolescent populations. Prospective cohort studies incorporating repeated dietary assessments are essential to more accurately examine the contribution of UPFs to habitual dietary patterns and the onset of mental health disorders, with a particular focus on clarifying the direction and temporal sequence of these associations. Although challenging to implement due to potential ethical and health concerns, intervention studies would offer more robust evidence regarding causal relationships. In parallel, mechanistic studies involving human participants are needed to identify the specific characteristics of UPFs that may negatively affect mental health and to determine whether the observed bidirectional associations reflect causal pathways. Future research should also prioritize the adoption of standardized methodologies to enhance comparability across studies and to quantify threshold levels of UPF intake that may be linked to adverse mental health outcomes. The development and validation of dedicated instruments to accurately measure UPF consumption would further advance the field. Moreover, investigating sex-specific variations in the relationship between UPF intake and mental health is an important area for future inquiry, given that our study exclusively examined adolescent girls and previous evidence indicates that both the prevalence of depression and its association with dietary patterns may differ by sex.

Conclusion

Our study demonstrates a significant positive association between the consumption of ultra-processed foods (UPFs) and elevated odds of depression, stress, and anxiety among adolescent girls, with the most pronounced association identified for depression. These findings contribute to the expanding body of literature that implicates UPF intake in adverse mental health outcomes and underscore the relevance of dietary considerations in strategies aimed at promoting and safeguarding mental health in adolescent populations. Efforts to reduce UPF consumption and encourage the adoption of healthier dietary patterns may offer meaningful benefits for both psychological and physical well-being within this at-risk group.

Authors’ contributions

Conceptualization, AJ and ZY; Formal analysis, ZY; Methodology, ANT, MM and VB; Project administration, ZY and AJ; Writing – original draft, AJ and ZY; Writing – review & editing, ZY. All authors read and approved.

Funding

No Funding.

Data availability

The datasets analyzed in the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the principles outlined in the Declaration of Helsinki and received ethical approval from the Ethics Committee of the National Nutrition and Food Technology Research Institute at Shahid Beheshti University of Medical Sciences, Tehran, Iran (ethics approval number 054577). Informed written consent was obtained from all participants prior to their inclusion in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.