Adherence to the EAT–Lancet planetary health diet and its associations with obesity and health-related quality of life in Iranian adolescents: a cross-sectional study

Abstract

Background

Adolescent obesity and diminished health-related quality of life (HRQoL) remain critical global issues and are further complicated by the urgent need for sustainable diets amid climate change. The Planetary Health Diet Index (PHDI), which is based on EAT–Lancet recommendations, connects human nutrition with environmental sustainability.

Methods

This cross-sectional study of 1,038 students (mean age, 15.13 ± 1.53 years) from the 2024–2025 academic year assessed dietary intake via a validated 147-item food frequency questionnaire. Anthropometrics followed standard protocols, and HRQoL was evaluated with the Pediatric Quality of Life Inventory™ Version 4.0 Generic Core Scales (PedsQL). Adjusted binary logistic regression models, controlling for age, sex, ethnicity, socioeconomic status, sleep duration, screen time, and energy intake, were used to analyze associations (SPSS version 28).

Results

The mean PHDI score was 50.79 (95% CI, 50.13–51.45; range, 19.66–94.46). According to the adjusted analyses, the highest quartile was associated with lower odds of general obesity (OR, 0.44; 95% CI, 0.24–0.81), abdominal obesity (OR, 0.56; 95% CI, 0.36–0.89), and impaired HRQoL (OR, 0.41; 95% CI, 0.24–0.70; all P < 0.05).

Conclusion

Greater PHDI adherence is associated with improved anthropometric status and HRQoL in adolescents, underscoring the potential of sustainable diets for holistic health promotion. Longitudinal studies are needed to establish causality and generalizability.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-025-26174-7.

Background

Human health, diet, and climate change are so deeply intertwined that we cannot afford to address them in isolation [1, 2]. Childhood and adolescent obesity are among the most pressing global public health challenges [3, 4], with prevalence rates particularly elevated in the Middle East and North Africa (MENA) region—where according to UNICEF, obesity among school-aged children and adolescents has alarmingly doubled since 2000 [5, 6]. In Iran, a 2024 meta-analysis reported rates of 12.43% for overweight and 6.51% for obesity among Iranian children and adolescents [7], a stark illustration of the regional burden of this multifaceted health challenge. The consequences of adolescent obesity extend beyond physical health issues, including cardiovascular risk factors and metabolic disorders, to encompass significant psychological impacts such as depression, anxiety, and social stigmatization that can persist into adulthood [8, 9]. With these far-reaching effects in mind, health-related quality of life (HRQoL), defined as “the physical, psychological, and social domains of health, seen as distinct areas that are influenced by a person’s experiences, beliefs, expectations, and perceptions,” has emerged as a validated framework for systematically capturing the quality-of-life impacts that traditional clinical measures fail to assess [10, 11].

With a focus on prevention, contemporary nutritional epidemiology increasingly highlights overall diet quality —rather than individual nutrients— as a modifiable correlate of obesity risk and psychosocial well-being in adolescents [12, 13]. While related research provides contextual insights—for instance, a systematic review by Romero-Robles and colleagues suggested a positive correlation between Mediterranean Diet (MD) adherence and HRQoL in children and adolescents [14]. Similarly, a study by Visser et al. in Brazilian adolescents found that lower fruit and vegetable intake or increased fast-food consumption was associated with lower HRQoL scores [15]. To the best of our knowledge, no studies to date have specifically examined these associations in relation to adherence to the EAT-Lancet “Planetary Health Diet” (PHD), a regimen that uniquely integrates planetary sustainability with human health in a globally applicable framework. This represents a key research gap, particularly among adolescents in diverse cultural contexts like Iran, where the potential of the PHD to simultaneously promote individual well-being and environmental health remains unexplored [16].

The planetary health diet index (PHDI) represents an innovative approach to dietary assessment that quantifies adherence to the EAT-Lancet Commission’s 2019 reference diet, simultaneously optimizing human health and environmental sustainability [17]. In the 2025 EAT-Lancet update, dietary intake recommendations remained unchanged, with the addition of “justice” as a core principle and an emphasis on local interpretation and adaptation of the universally applicable planetary health diet to reflect the culture, geography, and demography of populations and individuals [18]. The PHDI consists of 16 components, and the maximum achievable score is 150 with energy adjustment, emphasizes plant-forward, nutrient-dense foods while limiting the consumption of red meat, animal fats, added sugars, refined grains, and starchy vegetables [17, 19]. The convergence of sustainability and nutrient adequacy within the PHDI creates unique potential for dual benefits: many plant-based foods prioritized in sustainable dietary patterns—such as fruits, vegetables, legumes, nuts, and whole grains—are inherently rich in fiber, phytochemicals, and micronutrients that support healthy body weight regulation, metabolic function, and mental health as seen in adolescent cohorts from Project EAT [20–22]. Evidence from adult populations suggests that adherence to sustainable dietary patterns may support weight management, with Reger et al. reporting a 39% lower obesity risk [23] and Cacau et al.’s findings in Brazilian adults reductions in BMI and waist circumference [24]. In adolescents, however, research remains limited and mixed. For example, Murcia-Lesmes et al. observed lower BMI z-scores with higher PHDI adherence [25], whereas Cacau et al. reported no significant association with BMI despite improvements in other cardiometabolic markers. These inconsistencies, combined with the scarcity of adolescent studies, highlight the need for age-specific research to clarify these relationships [26].

Based on existing evidence, the current literature has limitations that hinder a comprehensive understanding of these associations. Globally, limited studies have examined the relationship between PHDI or planetary health diets and obesity status in adolescents, and most focusing on adult populations [23, 27]. To our knowledge, no direct investigations in Iran or the broader MENA have been conducted on adolescents to integrate PHDI with obesity, nor have they combined these with multidimensional, validated HRQoL assessment tools like the PedsQL 4.0 Generic Core Scales; this despite the region’s unique confluence of rising obesity prevalence, meat-centric dietary traditions, and environmental pressures such as water scarcity from climate change and food system vulnerabilities [28]. The present study addresses these knowledge gaps by evaluating these associations among adolescents in Mashhad, Iran, hypothesizing that higher PHDI adherence may be inversely related to obesity-related anthropometric indices and directly related to various HRQoL domains. Specifically, this cross-sectional study aims to estimate the association between adherence to the EAT–Lancet Planetary Health Diet and obesity and HRQoL, comparing adolescents with higher versus lower adherence levels.

Methods

Study population and sampling

This cross-sectional study enrolled 1038 students (aged 13–18 years) from middle (grades 7–9) and high (grades 10–12) schools in Mashhad, Iran, using multistage stratified cluster random sampling from October 2, 2024, to January 5, 2025, with approvals from the Khorasan Razavi Province Education Department. Mashhad’s seven educational districts were stratified by socioeconomic status (affluent, semi-affluent, deprived) [29], with one district randomly selected per stratum (districts 1, 4, and 7, respectively); within each, lists of girls’ and boys’ schools were obtained, and two middle and two high schools per gender were randomly selected using probability proportional to size. As shown in Fig. 1, a total of 1,112 students expressed willingness to participate in the study. Of these, 74 participants were excluded for the following reasons: diagnosed diseases or use of medications that could significantly affect physical/mental health or nutritional status (n = 13); weight control diets or ≥ 70 missing items in the Food Frequency Questionnaire (FFQ) (n = 15) [30]; daily energy intake outside the range of 800-4,200 kilocalories (n = 41) [31, 32]; and more than 50% missing data in the PedsQL Questionnaire (n = 5) [33]. Therefore, data from 1,038 eligible participants were analyzed. Notably, none of these participants reported a history of alcohol consumption or smoking.

Fig. 1.

Participant flow throughout the study

To further ensure data quality, missing data in the FFQ and other questionnaires were handled using complete-case analysis after these exclusions. For the remaining participants, any sporadic missing values were not imputed, as they were minimal (< 5% per participant), and sensitivity analyses confirmed no significant impact on results.

The Research Ethics Committee of Mashhad University of Medical Sciences, Mashhad, Iran, approved the study protocol (approval number IR.MUMS.MEDICAL.REC.1403.211). All procedures were performed in accordance with the ethical standards set forth in the 1964 Helsinki Declaration and its subsequent amendments. Parents and legal guardians of adolescents participating in this study were asked for passive informed consent and adolescents themselves were asked for active informed consent before the start of the study.

Measurements

All study measurements were obtained during a single assessment period.

Sociodemographic, anthropometric, and lifestyle characteristics

Data on age, sex, ethnicity, socioeconomic status, supplement use, sleep duration, and screen time were collected via self-reported questionnaire. Socioeconomic status was assessed using the Family Affluence Scale II (FAS-II), generating scores from 0 to 9 classified as low (0–2), medium [3–5], or high [6–9, 34].

Two trained nutrition specialists performed all anthropometric measurements, with gender-appropriate assignments. Weight was measured using a digital scale (Seca Instruments, Germany) with a precision of 100 g, while participants wore only light clothing and were barefoot. Height was measured using a stadiometer (accuracy 0.1 cm) while individuals stood with relaxed shoulders and without shoes. Body mass index (BMI) was determined by dividing an individual’s weight in kilograms by the square of their height in meters. To classify general obesity, we used age- and sex-specific BMI percentiles derived from a large Iranian population-based study (CASPIAN-IV), which provides culturally relevant references for Iranian children and adolescents [35]. Participants were classified as having general obesity if their BMI was ≥ 95th percentile for age and sex [35].

Waist circumference (WC) was assessed with participants standing upright, using a non-stretchable anthropometric tape positioned horizontally at the midpoint between the lower rib margin and the iliac crest at the end of a normal exhalation [36]. The two independent WC readings were averaged to obtain a mean value. For classifying abdominal obesity, we applied modified criteria specific to Iranian adolescents, defining it as WC ≥ 90th percentile for age and sex [35]. These percentiles were derived from smoothed age- and gender-specific charts using Cole’s LMS (lambda-mu-sigma) method [37], which normalizes data distributions and is commonly used for generating anthropometric reference values. This Iranian-specific reference was selected for its direct applicability to our study population, ensuring accurate identification of central adiposity risks in this demographic.

Dietary intakes

Data on participants’ usual food intake were collected using a validated 147-item FFQ [38, 39]. The validity of this FFQ had been previously tested among Iranian adolescents [38]. Participants recalled frequency and amount of food consumption over the past year through face-to-face interviews. Food items included standard serving sizes with consumption reported daily, weekly, monthly, or annually. All foods were converted to grams per day using household measures [40]. Nutrient intake values were then calculated for each participant via Nutritionist IV software (First Databank, San Bruno, CA, USA), which utilizes the USDA Food Composition Database supplemented with traditional Iranian food items where applicable [41, 42]. All nutrient values represent estimated intakes derived from FFQ data and do not reflect circulating biomarkers or laboratory measurements.

Calculation of the PHDI

The EAT-Lancet Commission on “Healthy Diets from Sustainable Food Systems” presented a dietary model known as the “Planetary Health Diet,” aimed at simultaneously promoting human health and environmental sustainability. The PHDI, was previously validated in an Iranian population by Shojaei et al. [43]. The PHDI was derived from recommendations presented in the EAT-Lancet Commission’s proposed reference diet [1]. The method used to calculate the PHDI was developed by Cacau et al. [44]. The PHDI consists of 16 components, with a maximum achievable score of 150. A higher PHDI score indicates greater adherence to the planetary health diet.

The PHDI used the recommended intake ranges and midpoints specified in the reference planetary health diet of 2,500 kilocalories per day [1]. Subsequently, these values were converted to the percentage of calories each food group contributes to the total diet. Food groups were categorized into adequacy, optimum, ratio, and moderation components according to the predefined PHDI framework, which operationalizes the EAT–Lancet Commission’s reference diet by integrating nutritional adequacy and environmental sustainability considerations derived from global evidence. Adequacy components include those foods where complete avoidance (i.e., zero consumption) tends to diminish overall diet quality, while intakes meeting or exceeding established reference levels generally present little risk to human health or the planet. This category encompasses nuts, legumes, fruits, all vegetables, and whole grains.

Optimum components, by contrast, are defined by the benefits of a moderate minimum intake—often aligned with midpoint values in reference diets—over none. Yet, as consumption approaches or exceeds an upper threshold drawn from these references, it may compromise both dietary balance and ecological viability. Foods in this group include eggs, fish and seafood, potatoes, dairy products, and unsaturated oils. The PHDI also incorporates two ratio-based components that assess the share of dark green vegetables and red and orange vegetables within total vegetable consumption. To avoid placing undue weight on this element, each ratio is capped at a maximum of 5 points. Finally, moderation components are those were reducing intake toward zero enhances both diet quality and sustainability outcomes. This applies to red meat, poultry and its alternatives, animal fats, and added sugars (Table 1).

Table 1.

Planetary health diet index components, standards for scoring (caloric densities), and corresponding point values

¹All values expressed as caloric densities from the reference diet proposed by the EAT-Lancet Commission. The bars represent the limits. £ Red meat: beef, lamb, and pork. ¢ Legumes: beans and soy. § Dairy: excluding dairy fats. º Unsaturated oils: including palm oil. /= DGV/total ratio: ratio between the energy intake of dark green vegetables (numerator) and the total of vegetables (denominator) multiplied by 10. ≡ ReV/total ratio: ratio between the energy intake of red and orange vegetables (numerator) and the total of vegetables (denominator) multiplied by 10. ǂ Animal fat: lard, tallow, and dairy fats. DGV/total ratio: dark green vegetable/total ratio. ReV/total ratio: red vegetable/total ratio

Health-related quality of life

The validated and reliable Persian version of the 23-item Pediatric Quality of Life Inventory™ Version 4.0 Generic Core Scales (PedsQL) for healthy adolescents aged 13–18 years was used to assess HRQoL [45, 46]. This brief, easy-to-complete, multidimensional self-report instrument is designed to measure core health dimensions as defined by the WHO, as well as school functioning over the past month [46, 47].

All 23 questions, answered on a 0–4 scale (i.e., never, almost never, sometimes, often, and almost always), are first reverse-scored and linearly transformed to a 0-100 scale as follows: never = 100; almost never = 75; sometimes = 50; often = 25; and almost always = 0 [12, 13]. PedsQL scale and summary scores, all ranging from zero to 100, are then calculated based on the following formulas: Total scale score = (sum of items 1–23) ÷ 23; Physical health summary score = (sum of items 1–8) ÷ 8; Psychosocial health summary score = (sum of items 9–23) ÷ 15; emotional functioning score = (sum of questions 9–13) ÷ 5; social functioning score = (sum of questions 14–18) ÷ 5; and school functioning score = (sum of questions 19–23) ÷ 5. Higher PedsQL scores indicate better health-related quality of life [12, 13]. Based on the recommendation of Varni et al. [48] impaired health-related quality of life was defined as more than 1 standard deviation below the mean total PedsQL scores of the entire sample population. Internal consistency reliability of the PedsQL was evaluated using Cronbach’s alpha coefficient in SPSS, with values ≥ 0.70 considered acceptable [49]. Subscale and total scale alphas were computed to confirm the instrument’s reliability in this sample.

Statistical analysis

The percentage of calories from the 16 EAT-Lancet food groups was calculated, followed by derivation of the PHDI score as previously described. Participants (total n = 1038) were stratified into PHDI quartiles (n ≈ 259–260 per quartile) for analysis. Continuous variables were reported as means ± standard deviations (SD), and categorical variables as numbers and percentages. Differences across quartiles were assessed using chi-square tests for qualitative variables and analysis of variance (ANOVA) for quantitative ones. Energy-adjusted nutrient intakes were compared via analysis of covariance (ANCOVA), with results as means ± standard errors (SE), to control for total energy intake.

Normality of continuous variables was evaluated using the Kolmogorov–Smirnov test and Q-Q plots; non-normal variables were log-transformed before analysis. Given that data were clustered within schools, multilevel null models (unconditional random intercept models, i.e., two-level linear mixed models) [50] were initially fitted for all Pediatric PedsQL scores and anthropometric measures to assess whether the level-2 clustering variable (i.e., School ID) significantly influenced the intercept (mean) of the level-1 dependent variables (i.e., individual-level outcomes). As no significant clustering effects were observed in these null models, multilevel modeling was deemed unnecessary, and standard parametric tests were proceeded with. Multivariable-adjusted means of PedsQL scores and anthropometric measures were computed and compared across PHDI quartiles using one-way ANCOVA. Binary logistic regression yielded crude and multivariable-adjusted odds ratios (ORs) with 95% confidence intervals (CIs) for obesity and impaired HRQoL by PHDI quartile, and for impaired HRQoL per one-unit increase in BMI-for-age z-score. Multivariable models adjusted for age, sex, ethnicity, socioeconomic status, sleep duration, screen time, and energy intake (kcal/day) as confounders. Potential confounders were selected based on prior literature [51, 52], univariate associations with PHDI and outcomes (p < 0.05), and a change-in-estimate approach (> 10% alteration in the PHDI coefficient upon adjustment). Missing data were handled using complete-case analysis after exclusions for excessive missing FFQ items. To further assess dose–response relationships, the PHDI score was additionally modeled as a continuous variable standardized to one standard deviation, and odds ratios were estimated per 1-SD increase in PHDI. Sensitivity analyses included re-running regressions with alternative energy adjustment cutoffs (± 300 kcal around the primary range of 800–4,200 kcal) [30, 32], yielding consistent results. Linear trends were tested by modeling quartile medians as continuous predictors in regression analyses. Pearson correlations assessed associations between PHDI and PedsQL scores. All analyses were performed using SPSS Statistics version 28 (IBM Corp., Armonk, NY, USA). A two-tailed p-value < 0.05 was considered statistically significant.

Results

Demographic, socioeconomic, and anthropometric characteristics

A total of 1,038 adolescents participated in this study (Fig. 1). Participant characteristics are summarized in Table 2. The mean age of participants was 15.13 ± 1.53 years, with 43.0% male and 57.0% female students. Regarding socioeconomic status, 17.8% of participants were classified as having low, 56.0% as middle, and 26.2% as high socioeconomic status. Anthropometric measurements showed a mean BMI z-score of 0.31 ± 1.33 and a mean waist circumference z-score of 0.13 ± 0.95. General obesity was observed in 11.7% of participants, while abdominal obesity showed higher prevalence at 21.6%. The average daily sleep duration was 7.85 ± 1.76 h, and the mean daily television viewing time was 3.22 ± 2.14 h (Table 2).

Table 2.

Characteristics of study participants by quartiles(Q) of the PHDI (N=1038)^a,b,c

Characteristics	Total (1038)	planetary health diet index scores				P-value
		Q1 <42.95 (n=259)	42.95<Q2<49.85 (n=260)	49.85<Q3<57.46 (n=260)	Q4 >57.46 (n=259)
Mean Age (years, SD)	15.13 ± 1.53	15.41 ± 1.50	15.14 ± 1.55	15.01 ± 1.50	14.97 ± 1.51	0.004
Gender						0.014
Male	446 (43.0%)	122 (47.1%)	126 (48.5%)	104 (40.0%)	94 (36.3%)
Female	592 (57.0%)	137 (52.9%)	134 (51.5%)	156 (60.0%)	165 (63.7%)
Ethnicity						0.141
Persian	968 (93.3%)	240 (92.7%)	237 (91.2%)	250 (96.2%)	241 (93.1%)
Other	70 (6.7%)	19 (7.3%)	23 (8.8%)	10 (3.8%)	18 (6.9%)
SES						0.099
Low	185 (17.8%)	54 (20.8%)	43 (16.5%)	41 (15.8%)	47 (18.1%)
Medium	581 (56.0%)	147 (56.8%)	133 (51.2%)	147 (56.5%)	154 (59.5%)
High	272 (26.2%)	58 (22.4%)	84 (32.3%)	72 (27.7%)	58 (22.4%)
Average Sleep Duration (hours/day)	7.85 ± 1.76	7.69 ± 1.76	7.87 ± 1.76	7.97 ± 1.92	7.85 ± 1.58	0.325
Time Spent Watching TV (hours/day)	3.22 ± 2.14	3.38 ± 2.43	3.26 ± 2.26	3.18 ± 1.90	3.09 ± 1.91	0.462
Supplement intake	206 (19.8%)	49 (18.9%)	48 (18.4%)	53 (20.3%)	56 (21.6%)	0.798
PHDI scores	50.79 ± 10.82	38.07 ± 4.26	46.56 ± 2.05	53.45 ± 2.12	65.18 ± 6.99	<0.001
PedsQL scoresTotal	77.98 ± 13.91	75.77 ± 14.33	76.85 ± 13.86	77.66 ± 13.65	81.67 ± 13.15	<0.001
Physical Health	80.89± 15.55	79.27 ± 15.95	79.86 ± 15.53	80.29 ± 15.54	84.16 ± 14.75	0.001
Psychosocial Health	76.40 ± 15.30	73.76 ± 16.11	75.26 ± 15.15	76.26 ± 14.82	80.34 ± 14.40	<0.001
Emotional Functioning	68.74 ± 22.29	64.9 ± 22.81	68.00 ± 22.10	68.31 ± 22.08	73.74 ± 20.49	<0.001
Social Functioning	81.05 ± 17.19	79.52 ± 18.22	79.64 ± 17.31	80.60 ± 17.31	84.48 ± 15.40	0.002
School Functioning	79.06 ± 16.17	76.31 ± 17.02	77.27 ± 15.83	79.87 ± 15.46	82.81 ± 15.63	<0.001
BMI (z-score)	0.31 ± 1.33	0.45 ± 1.32	0.33 ± 1.44	0.32 ± 1.30	0.12 ± 1.27	0.042
WC (z-score)	0.13 ± 0.95	0.29 ± 0.99	0.18 ± 1.02	0.10 ± 0.89	-0.04 ± 0.86	0.001
General Obesity	122 (11.7%)	40 (15.4%)	38 (14.6%)	29 (11.1%)	15 (5.7%)	0.006
Abdominal Obesity	224 (21.6%)	65 (25.1%)	63 (24.2%)	55 (21.1%)	41 (15.8%)	0.008

PHDI planetary health diet index, PedsQL the Pediatric Quality of Life Inventory™ Version 4.0 Generic Core Scales, HRQoL health-related quality of life, SES socioeconomic status, BMI body mass index, WC Waist Circumference

^aData are presented as mean ± SD or n (%), except for BMI and Waist Circumference Z-score which is reported as arithmetic mean

^bThe chi-square test and the one-way analysis of variance were used for comparison of categorical and continuous variables among the quartiles of dietary pattern scores, respectively

^cImpaired HRQoL was defined as > 1 standard deviation below the total population sample mean PedsQL scores

PHDI scores and distribution

The mean PHDI score among all participants was 50.79 (95% CI: 50.13–51.45), with scores ranging from 19.66 to 94.46 (Table 2). When participants were categorized by PHDI quartiles, the proportion of female students was higher progressively across quartiles, being highest in the fourth quartile (63.7% compared with 52.9% in Q1, P = 0.014). Participants in the highest PHDI quartile were significantly younger (14.97 ± 1.51 compared with 15.41 ± 1.50 years, P = 0.004) and showed more favorable anthropometric profiles, including significantly lower BMI z-scores (0.12 ± 1.27 compared with 0.45 ± 1.32, P = 0.042), waist circumference z-scores (− 0.04 ± 0.86 compared with 0.29 ± 0.99, P = 0.001), and lower prevalence of both general obesity (5.7% compared with 15.4%, P = 0.006) and abdominal obesity (15.8% compared with 25.1%, P = 0.008) (Table 2).

HRQoL assessment

The internal consistency of the PedsQL was assessed using Cronbach’s alpha, with values ≥ 0.70 considered acceptable [49]. The total score showed excellent reliability (α = 0.891), and all subscales were acceptable (see Supplementary Table 1 for subscale-specific values and confidence intervals). The overall prevalence of impaired HRQoL based on total, physical health, psychosocial health, emotional functioning, social functioning, and school functioning PedsQL scores was 16.6%, 17.1%, 16.4%, 16.4%, 13.6%, and 14.9%, respectively. Compared to those in the bottom quartiles of PHDI score, participants in the top quartile had higher PedsQL scores, and were less likely to have impaired HRQoL (all P ≤ 0.05) (Table 2).

Dietary intake analysis

Mean dietary intake patterns among study participants are presented in Table 3. The mean daily energy intake was 2,556 kcal, with macronutrient distribution of 14.5% energy from protein, 56.7% from carbohydrates, and 28.7% from fat.

Table 3.

Mean dietary intake variables by quartiles (Q) of the PHDI (n=1038)^a,b

Variable	Total (1038)	planetary health diet index scores				P-value
		Q1 <42.95 (n=259)	42.95 <Q2<49.85 (n=260)	49.85 <Q3<57.46 (n=260)	Q4 >57.46 (n=259)
Energy intake (kcal/day)	2555.95 ± 635.14	2640.34 ± 630.39	2570.51 ± 620.79	2568.69 ± 654.69	2444.16 ± 618.39	0.01
Energy from protein (%)	14.54 ± 2.35	14.97 ± 2.32	14.78 ± 2.35	14.25 ± 2.45	14.15 ± 2.19	<0.001
Energy from carbohydrate (%)	56.70 ± 6.43	53.77 ± 5.89	56.04 ± 5.58	57.13 ± 6.39	59.84 ± 6.33	<0.001
Energy from fat (%)	28.75 ± 6.95	31.25 ± 6.58	29.16 ± 6.10	28.60 ± 6.20	25.99 ± 7.06	<0.001
Daily nutrient intake per 1000 kcal
Protein (g)	36.35 ± 5.89	37.42 ± 5.80	36.96 ± 5.89	35.64 ± 6.14	35.39 ± 5.49	<0.001
Carbohydrate (g)	141.76 ± 16.08	134.43 ± 14.72	140.12 ± 13.97	142.84 ± 15.97	149.61 ± 15.84	<0.001
Fiber (g)	15.73 ± 4.52	13.82 ± 3.80	14.98 ± 3.98	15.99 ± 4.07	18.13 ± 5.02	<0.001
Fat (g)	31.94 ± 7.73	34.72 ± 7.31	32.40 ± 6.78	31.78 ± 7.82	28.88 ± 7.85	<0.001
Saturated fatty acid (g)	11.97 ± 3.09	13.41 ± 2.80	12.47 ± 2.95	11.56 ± 2.86	10.45 ± 2.95	<0.001
Monounsaturated fatty acid (g)	12.51 ± 3.02	13.14 ± 3.07	12.43 ± 2.66	12.57 ± 3.18	11.89 ± 3.03	<0.001
Cholesterol (mg)	114.98 ± 55.24	143.59 ± 57.00	125.88 ± 53.70	107.18 ± 51.42	83.25 ± 38.14	<0.001
Vitamin A (μg)	267.83 ± 154.07	268.66 ± 169.90	281.44 ± 152.73	254.87 ± 167.14	266.35 ± 121.36	0.273
Beta-carotene (μg)	1489.36 ± 1231.23	1255.31 ± 1222.49	1463.68 ± 1276.70	1428.77 ± 1224.22	1810.00 ± 1138.09	<0.001
Vitamin D (μg)	0.92 ±0.61	1.07 ± 0.67	1.04 ± 0.64	0.82 ± 0.54	0.74 ± 0.51	<0.001
Vitamin E (mg)	5.41 ± 2.28	4.90 ± 1.94	5.04 ± 1.88	5.80 ± 2.78	5.89 ± 2.24	<0.001
Vitamin K (μg)	91.33 ± 81.43	78.10 ± 73.45	86.57 ± 64.73	87.31 ± 99.20	113.38 ± 80.44	<0.001
Vitamin B1 (mg)	0.82 ± 0.15	0.78 ± 0.13	0.80 ± 0.14	0.83 ± 0.15	0.86 ± 0.15	<0.001
Vitamin B2 (mg)	0.83 ± 0.19	0.89 ± 0.19	0.87 ± 0.19	0.80 ± 0.17	0.76 ± 0.16	<0.001
Daily nutrient intake per 1000 kcal
Vitamin B3 (mg)	9.67 ± 1.98	9.77 ± 2.09	9.66 ± 2.00	9.54 ± 1.91	9.72 ± 1.91	0.574
Vitamin B5 (mg)	2.33 ± 0.42	2.35 ± 0.46	2.38 ± 0.42	2.28 ± 0.41	2.30 ± 0.38	0.026
Vitamin B6 (mg)	0.77 ± 0.13	0.76 ± 0.13	0.76 ± 0.13	0.77 ± 0.13	0.79 ± 0.13	0.022
Biotin (μg)	14.30 ± 3.89	14.34 ± 3.80	14.36 ± 3.84	14.01 ± 3.80	14.50 ± 4.10	0.522
Folate (μg)	237.41 ± 46.99	221.25 ± 38.40	230.83 ± 45.30	239.49 ± 46.74	258.07 ± 49.09	<0.001
Vitamin B12 (μg)	1.84 ± 1.07	2.13 ± 1.12	2.05 ± 1.05	1.78 ± 1.17	1.40 ± 0.75	<0.001
Vitamin C (mg)	47.61 ± 21.19	40.03 ± 19.00	47.50 ± 24.42	48.57 ± 24.30	54.38 ± 26.31	<0.001
Calcium (mg)	409.74 ± 109.24	433.80 ± 106.78	426.58 ± 113.68	394.09 ± 106.08	384.50 ± 101.71	<0.001
Phosphorus (mg)	646.91 ± 102.44	656.86 ± 101.27	658.35 ± 102.69	634.85 ± 103.70	637.59 ± 100.38	0.009
Magnesium (mg)	181.92 ± 34.20	170.13 ± 28.92	177.53 ± 29.99	180.88 ± 32.41	199.17 ± 38.07	<0.001
Iron (mg)	7.49 ± 1.27	6.97 ± 1.03	7.28 ± 1.17	7.55 ± 1.22	8.15 ± 1.36	<0.001
Zinc (mg)	5.23 ± 0.86	5.18 ± 0.83	5.24 ± 0.85	5.17 ± 0.91	5.32 ± 0.87	0.207
Copper (mg)	0.81 ± 0.16	0.75 ± 0.13	0.79 ± 0.14	0.81 ± 0.15	0.89 ± 0.16	<0.001
Manganese (mg)	2.83 ± 0.94	2.57 ± 0.77	2.68 ± 0.90	2.86 ± 0.97	3.22 ± 1.00	<0.001
Selenium (μg)	52.68 ± 13.99	50.99 ± 11.86	52.38 ± 13.75	52.88 ± 14.81	54.49 ± 15.15	<0.001
Potassium (g)	1.52 ± 0.32	1.46 ± 0.28	1.51 ± 0.30	1.51 ± 0.33	1.58 ± 0.35	<0.001
Sodium (g)	1.79 ± 1.11	1.87 ± 1.76	1.74 ± 0.77	1.76 ± 0.74	1.80 ± 0.85	0.557

PHDI planetary health diet index

^aAll values are reported as mean ± standard deviation

^bAll P-values were calculated using one-way analysis of variance

Dietary analysis by PHDI quartiles showed significant differences in nutrient intake patterns using one-way ANOVA. Participants in the highest quartile were associated with more favorable estimated nutrient intake profiles, including higher intakes of carbohydrates, fiber, beta-carotene, vitamins E, C, K, B1, B6, folate, magnesium, iron, copper, manganese, selenium, and potassium compared to those in the lowest quartile (all P < 0.001). Conversely, those in the highest PHDI quartile had significantly lower intakes of total energy, fat, saturated fatty acids, monounsaturated fatty acids, cholesterol, protein, vitamins D, B2, B12, calcium, and phosphorus compared to the lowest quartile (all P < 0.05) (Table 3).

Associations between PHDI and health outcomes

Table 4 presents the associations between PHDI quartiles and odds of obesity and impaired HRQoL. Following adjustment for potential confounding variables (age, sex, ethnicity, socioeconomic status, sleep duration, screen time, and energy intake) in multivariable binary logistic regression analyses, participants in the highest PHDI quartile of the planetary health diet, compared with those in the lowest quartile, demonstrated significantly lower odds of both general obesity (OR: 0.44, 95% CI: 0.24–0.81, P < 0.001) and abdominal obesity (OR: 0.56, 95% CI: 0.36–0.89, P < 0.001). Additionally, the highest quartile showed a significant lower mean BMI z-score of − 0.33 units (95% CI: −0.55, − 0.11, P < 0.05). In continuous analyses, each one–standard deviation increase in PHDI score was associated with 27% lower odds of general obesity (OR = 0.73; 95% CI: 0.59–0.91) and 19% lower odds of abdominal obesity (OR = 0.81; 95% CI: 0.69–0.95) (Supplementary Table 2).

Table 4.

Risk of having general and abdominal obesity and impaired HRQOL by quartiles (Q) of PHDI (n = 1038)^a,b,c,d

PedsQL scores	Planetary health diet index scores
	Q1 <42.95 (n=259)	42.95 <Q2<49.85 (n=260)	49.85 <Q3<57.46 (n=260)	Q4 >57.46 (n=259)	P-trend
Physical Health Disorder
Model 1	1.00	0.78 [0.50-1.23]	0.57 [0.36-0.93]	0.38 [0.22-0.65]	<0.001
Model 2	1.00	0.83 [0.53-1.32]	0.60 [0.37-0.97]	0.37 [0.21-0.63]	<0.001
Psychosocial Health Disorder
Model 1	1.00	0.80 [0.52-1.23]	0.57 [0.36-0.90]	0.48 [0.29-0.77]	<0.001
Model 2	1.00	0.82 [0.52-1.29]	0.63 [0.39-1.01]	0.53 [0.32-0.87]	<0.001
Emotional Function Disorder
Model 1	1.00	0.82 [0.53-1.26]	0.69 [0.44-1.08]	0.47 [0.29-0.76]	0.002
Model 2	1.00	0.88 [0.55-1.35]	0.65 [0.41-1.06]	0.55 [0.33-0.90]	<0.001
Social Function Disorder
Model 1	1.00	0.84 [0.52-1.33]	0.72 [0.47-1.16]	0.45 [0.26-0.77]	0.003
Model 2	1.00	0.85 [0.52-1.36]	0.79 [0.48-1.29]	0.47 [0.27-0.81]	<0.001
Academic Function Disorder
Model 1	1.00	0.97 [0.62-1.50]	0.54 [0.33-0.88]	0.46 [0.28-0.77]	<0.001
Model 2	1.00	1.02 [0.65-1.61]	0.63 [0.38-1.04]	0.51 [0.30-0.86]	<0.001
HRQOL Disorder
Model 1	1.00	0.83 [0.54-1.29]	0.63 [0.40-0.98]	0.42 [0.25-0.71]	<0.001
Model 2	1.00	0.89 [0.57-1.40]	0.68 [0.42-1.10]	0.41 [0.24-0.70]	<0.001
General Obesity
Model 1	1.00	1.06 [0.64-1.73]	0.77 [0.46-1.31]	0.46 [0.25-0.83]	0.006
Model 2	1.00	0.98 [0.59-1.64]	0.73 [0.43-1.26]	0.44 [0.24-0.81]	<0.001
Abdominal Obesity
Model 1	1.00	1.01 [0.68-1.51]	0.78 [0.51-1.17]	0.55 [0.35-0.86]	0.004
Model 2	1.00	0.95 [0.63-1.44]	0.78 [0.51-1.20]	0.56 [0.36-0.89]	<0.001

PHDI planetary health diet index, HRQOL health-related quality of life, SD standard deviation, PedsQL the Pediatric Quality of Life Inventory™ Version 4.0 Generic Core ScalesData are presented as odds ratio (OR) (95% confidence interval [CI])

^aData are presented as odds ratio (OR) (95% confidence interval [CI])

^bCrude and multivariable-adjusted ORs and 95% CIs for impaired HRQoL across the quartiles of dietary pattern scores were computed using binary logistic regression analysis

^cModel 2 adjustments were made for age, sex, ethnicity, socioeconomic status, sleep duration, screen time, and energy intakeData are presented as odds ratio (OR) (95% confidence interval [CI])

^dImpaired HRQoL was defined as >1 standard deviation below the total population sample mean PedsQL scoresData are presented as odds ratio (OR) (95% confidence interval [CI])

Furthermore, participants in the highest quartile of PHDI adherence showed significantly lower odds of experiencing impaired HRQoL across all assessed domains compared to those in the lowest quartile (Table 4). Specifically, individuals with higher PHDI scores were linked to lower likelihood of impaired overall HRQoL (OR: 0.41, 95% CI: 0.24–0.70, P < 0.001), physical health (OR: 0.37, 95% CI: 0.21–0.63, P < 0.001), psychosocial health (OR: 0.53, 95% CI: 0.32–0.87, P < 0.001), emotional functioning (OR: 0.55, 95% CI: 0.33–0.90, P < 0.001), social functioning (OR: 0.47, 95% CI: 0.27–0.81, P < 0.001), and school functioning (OR: 0.51, 95% CI: 0.30–0.86, P < 0.001) (Table 4). Each one–standard deviation increase in PHDI score was also associated with 24% lower odds of impaired overall HRQoL (OR = 0.76; 95% CI: 0.63–0.91) (Supplementary Table 2). After adjustment for potential confounders, there was a 32% increase in the odds of impaired HRQoL per unit increase in BMI-for-age z-score (OR: 1.32; 95% CI: 1.15–1.51, P < 0.001) (Table 5). Sensitivity analyses with varied energy intake cutoffs (± 400 kcal around 800–4,200 kcal) yielded similar results, confirming robustness.

Table 5.

Risk of impaired HRQOL per one-unit increase in BMI-for-age z-score^a,b,c,d

Variable	OR [95% CI]	P-value
BMI-for-age z-score
Model 1	1.20 [1.06–1.37]	0.004
Model 2	1.32 [1.15–1.51]	<0.001

^aData are presented as odds ratio (OR) (95% confidence interval [CI])

^bCrude and multivariable-adjusted ORs and 95% CIs for impaired HRQoL across the quartiles of dietary pattern scores were computed using binary logistic regression analysis

^cModel 2 adjustments were made for age, sex, ethnicity, socioeconomic status, sleep duration, screen time, and energy intake

^dImpaired HRQoL was defined as >1 standard deviation below the total population sample mean PedsQL scores

Discussion

This study examined associations between the PHDI and anthropometric status and health-related quality of life in Iranian adolescents and revealed a mean PHDI score of 50.79 (95% CI: 50.13–51.45) among participants. The mean PHDI score in the present study aligns with previous research on adolescents. Cacau et al. [53] reported a mean score of 44.3 among European adolescents in the HELENA study, whereas Marchioni et al. [19] reported a mean score of 43.6 among Brazilian adolescents from the 2017–2018 National Dietary Survey (n = 46,164). Similarly, Vargas-Quesada et al. [54] reported a mean PHDI score of 42.4 across 19,601 adolescents from six Latin American countries.

Our population may reflect regional dietary patterns and socioeconomic factors influencing food accessibility. Global adherence to the EAT-Lancet recommendations varies substantially by geography and economic status. Current fruit and vegetable consumption accounts for 57% of international recommendations globally (400 g per capita per day) [55, 56], ranging from 25% in parts of South Asia and Sub-Saharan Africa to 95% in select Mediterranean, Middle Eastern, and North African countries, including Armenia, Turkey, Tunisia, Romania, Egypt, and Iran [1, 57, 58]. Additionally, the prominent role of parents in shaping food choices among Iranian adolescents likely contributes to adherence, as studies indicate that parental modeling of healthy eating, the provision of nutritious home meals, and a cultural emphasis on family dining foster greater adherence to balanced diets, leading to healthier selections such as increased fruit and vegetable intake [59–61].

Despite these comparative results, our findings indicate substantial room for improvement in adolescent dietary quality. The participants achieved approximately one-third of the maximum PHDI score (0–150 scale), and adherence to healthy and sustainable dietary recommendations remains suboptimal. Lower scores likely are associated with higher consumption of red meat and processed foods, compounded by inadequate intake of nuts and whole grains observed in our population. These patterns mirror broader adolescent dietary trends, which are consistent with Akbari and Azadbakht’s [62] findings of inadequate whole grain, fruit, vegetable, and dairy consumption among Iranian adolescents, alongside excessive red meat, processed foods, and soft drink intake.

As a secondary descriptive analysis, Higher PHDI scores were associated with more favorable estimated nutrient intake profiles, including higher intakes of fiber, vitamins (B1, B6, C, E, K), and minerals (magnesium, iron, copper, manganese, selenium, potassium), and lower intakes of total energy and saturated fat. These patterns are consistent with the plant-forward composition of the EAT–Lancet reference diet [53] and serve primarily to confirm internal coherence of the PHDI construct rather than to imply biological effects. Importantly, nutrient intakes were estimated from dietary data and should not be interpreted as biomarkers of nutritional status.

Our findings showed a significant association between higher PHDI scores and lower obesity odds in adolescents, with those in the highest quartile showing 56% lower odds of general obesity and 44% lower odds of abdominal obesity than those in the lowest quartile. These results highlight the potential of sustainable, plant-rich diets to be associated with healthy anthropometric profiles during critical developmental stages. Based on these findings, suggestions to support adolescent diet quality in Iran include implementing school-based nutrition education programs emphasizing plant-rich foods, policy measures to subsidize affordable nuts and whole grains, and community health campaigns integrating sustainability principles; these could inform broader policies, enhance health education, and be associated with adolescent nutrition nationwide [63, 64].

Evidence linking adherence to EAT-Lancet–based dietary indices with adolescent health is still limited, and, importantly, no studies have been conducted in the MENA region. Existing research—mainly from European cohorts—generally supports our findings. In Spain, Murcia-Lesmes et al. found that higher PHDI scores were associated with lower BMI z-scores and better cardiometabolic markers [25]. A multicounty European analysis by Cacau et al. employing the PHDI found no significant association with BMI while other cardiometabolic markers improved [26]. In Spain, an intervention by Ojeda-Rodríguez et al. using the EAT-Lancet Diet Score demonstrated that increasing adherence reduced BMI, waist circumference, and metabolic risk factors [27].

Several mechanisms may explain these associations. Plant-based diets, with their favorable macronutrient composition, high fiber content, and low energy density, likely are linked to weight management through multiple pathways. The PHDI’s emphasis on reduced simple sugar consumption is correlated with better diet quality, as sugar-sweetened beverages are linked to weight gain through high-calorie, low-nutrient content and disrupted satiety signaling [65–67]. Higher fiber intake, emphasized in sustainable diets, is associated with reduced waist circumference through decreased energy density, enhanced satiety via intestinal hormone stimulation (CCK1, GLP-12), and slower gastric emptying [68, 69]. Reduced red meat consumption may decrease visceral fat accumulation by limiting saturated fatty acid intake [70]. Replacing saturated fats with unsaturated fats from sources like nuts reduces fat accumulation while promoting satiety and thermogenesis [71–73]. Additionally, minimally processed foods in sustainable diets reduce exposure to endocrine-disrupting pesticides that interfere with metabolic hormones [74–76].

Another key finding is that in this population of adolescents, higher adherence to EAT-Lancet recommendations, reflected in higher PHDI scores, was associated with more favorable HRQoL among adolescents. Participants in the highest quartile of PHDI adherence had 59% lower odds of HRQoL impairment compared to those in the lowest quartile. This finding is novel, as evidence on the direct link between this dietary pattern and adolescent quality of life is limited. Beyond dietary patterns, higher BMI-for-age z-score was independently associated with poorer HRQoL, consistent with previous studies showing that excess adiposity can negatively affect physical comfort, self-esteem, social functioning, and emotional well-being [77]. Although BMI-for-age z-score was independently associated with poorer HRQoL, it was not included in primary PHDI–HRQoL models to avoid overadjustment, as it may lie on the causal pathway. Sensitivity analyses additionally adjusting for BMI-for-age z-score yielded similar results, suggesting that dietary quality may influence HRQoL through pathways beyond weight status. Future studies using formal mediation approaches are warranted.

Health-related quality of life is a multidimensional indicator encompassing physical, psychological, and social aspects of health [78]. The better scores in this indicator among individuals following the planetary health diet may may relate to multiple factors. In the PHDI, legumes, fruits, vegetables, and whole grains are classified in the adequacy group, with higher consumption of these foods corresponding to greater dietary sustainability [17]. Findings from systematic reviews have shown that dietary patterns with high intake of vegetables, fruits, nuts, and whole grains and low intake of sweets and processed meat play a protective role against physical and mental health disorders by providing essential bioactive compounds that reduce oxidative stress, inflammation, and support gut microbiota balance [79–81]. Results from a systematic review and meta-analysis of 17 observational studies examining associations between various diet quality indices and HRQOL among children and adolescents by Wu et al. [80] are consistent with our findings, showing that greater adherence to healthy diets characterized by high consumption of fruits, vegetables, and whole grains is associated with better overall HRQOL and its sub-dimensions.

Several limitations should be acknowledged when interpreting our findings; however, we moderated their impact through a large sample size and validated tools. One key limitation is the cross-sectional design, which precludes causal inference but provides a comprehensive snapshot in a large, representative sample of healthy Iranian adolescents, supporting hypothesis generation for future longitudinal studies. Additionally, excluding participants with trauma, diseases, or relevant medications limits generalizability to healthy Iranian adolescents only. This enhances internal validity by reducing extraneous influences on dietary patterns and HRQoL, though replication in diverse groups is needed. Physical activity was not directly assessed due to the lack of a validated age- and sex-appropriate questionnaire for this population. We adjusted for screen time and sleep duration, which are commonly used behavioral proxies for activity patterns in adolescents and may partially capture lifestyle factors associated with diet and health outcomes. Nonetheless, residual confounding by physical activity cannot be ruled out. We also adjusted for confounders like age, sex, socioeconomic status, and anthropometrics via multivariable logistic regression informed by literature. Residual confounding remains possible, particularly from unmeasured factors such as psychosocial stress, household food security, and parental dietary habits, which may influence both diet quality and HRQoL. Sensitivity analyses (e.g., E-values) indicate that unmeasured factors would need strong associations to nullify results, but this risk persists. Furthermore, no participants self-reported alcohol or smoking, unlike meta-analysis rates of 14.7% and 16.8% in Iranian adolescents [82], likely due to social desirability bias [83]—especially in culturally sensitive contexts. These behaviors link to poorer HRQoL [84, 85] and diets [86, 87], potentially overestimating our inverse planetary health diet-HRQoL associations. We controlled for other confounders, but objective measures are warranted. The use of validated food frequency questionnaires for diet assessment may introduce recall bias and measurement error, potentially attenuating associations. Multiple tests raise Type I error risk, though we focused on hypothesis-driven analyses. Cultural factors (e.g., religious dietary norms) may add subtle biases, calling for cross-cultural validation. Although we endeavored to balance these limitations through a substantial sample size and the utilization of validated instruments, longitudinal replication in diverse cohorts is imperative to corroborate our findings.

Conclusions

Our findings suggest that higher PHDI scores, characterized by increased consumption of vegetables, fruits, legumes, nuts, and whole grains with reduced intake of red meat and processed foods, are associated with more favorable anthropometric status and HRQoL among Iranian adolescents. However, the causality of these associations needs to be confirmed by prospective studies of high methodological rigor. It will also be interesting to see if our results could be replicated among different adolescent populations and through longitudinal designs that can better clarify temporal relationships between sustainable dietary patterns and health outcomes in this age group.

Abbreviations

BMI

Body mass index

CCK

Cholecystokinin

FAS-II

Family Affluence Scale II

FFQ

Food frequency questionnaire

GLP-1

Glucagon-like peptide-1

HRQoL

Health-related quality of life

MENA

Middle East and North Africa

PedsQL

Pediatric Quality of Life Inventory™

PHDI

Planetary Health Diet Index

SPSS

Statistical Package for the Social Science

Authors' contributions

SD, ZN and SRS designed research; SD and ZN conducted research; SD and ZD analyzed data; SD, SM and SRS wrote the paper. SRS and ZD had primary responsibility for final content. All authors read and approved the final manuscript.

Funding

This study was financially supported by Mashhad University of Medical Sciences, Mashhad, Iran (Project Number: 4022488).

Data availability

The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The Research Ethics Committee of Mashhad University of Medical Sciences, Mashhad, Iran, approved the study protocol (approval number IR.MUMS.MEDICAL.REC.1403.211). All procedures were performed in accordance with the ethical standards set forth in the 1964 Helsinki Declaration and its subsequent amendments. Parents and legal guardians of adolescents participating in this study were asked for passive informed consent and adolescents themselves were asked for active informed consent before the start of the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.