Food consumption impacts both human health and environmental resources, making the adoption of sustainable consumption practices essential for preserving future food production (1). Agriculture accounts for approximately 25% of total greenhouse gas emissions, occupies around 40% of the Earth’s surface, and consumes 70% of global freshwater resources (2). Thus, the global food system is surpassing several planetary boundaries, with its stability increasingly threatened by ecosystem overexploitation and pollution. Dietary changes aimed at fostering sustainable eating habits can significantly reduce the demand for food items with a high carbon footprint, which pose a threat to the environment (1, 2). Those foods include red meat, processed meat, ultra-processed foods, sugar, and refined grains, all of which offer low nutritional benefits, due to their high content of saturated fats, cholesterol, sodium, added sugars, and refined starches (2).

The Planetary Diet, a healthy and sustainable diet, aligns with the achievement of Sustainable Development Goals by encouraging the consumption of nutrient-rich foods such as vegetables, fruits, whole grains, legumes, nuts, and unsaturated fats (3). This dietary pattern prioritizes the inclusion of plant-based foods, which are rich sources of dietary fiber, antioxidant bioactive compounds, including (poly) phenols and carotenoids, and vitamins (provitamin A, C, E); all of them known for health-promoting properties (4, 5). Numerous studies in adults have explored the adherence to the Planetary Health Diet and its association with cardiovascular diseases (6–10), cardiovascular events (11), and mortality (8, 12–14), demonstrating promising benefits across these areas.

Most existing studies have focused on adult populations, limiting our understanding of the nutritional benefits of adopting the Planetary Health Diet at a younger age. Elevated biomarkers of cardiometabolic risk, such as lipid profile, blood pressure (BP), and plasma glucose, are known factors for cardiovascular disease in adolescents (15, 16) and cardiovascular events later in life (17). Therefore, the aim of this prospective cohort study is to assess the association between adherence to the Planetary Health Diet, measured by the Planetary Health Dietary Index (PHDI), and the risk of new-onset high blood pressure, obesity, and other elevated cardiometabolic risk factors during 4 years of follow-up in adolescents in Spain. Thus, we studied risk factors including obesity, high plasma glucose, low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), total cholesterol, high-density lipoprotein cholesterol (HDL-C), and non-HDL-C. Additionally, we examined the relationship between changes in these cardiometabolic parameters, and systolic and diastolic blood pressures (SBP and DBP), body mass index (BMI) z-score, and waist-to-height ratio (WHtR).

Study population characteristics, lifestyle, food consumption, and nutrient intake according to the adherence to PHDI are described in Tables 1, 2. At baseline, approximately half of the participants were female (49.1%), aged 12.0 years, with a maternal migrant background of approximately 20%, and a mean BMI of 20.3 kg/m². Out of a maximum of 140 points, adolescents had a mean score of 89.9 ± 13.1, with minimum and maximum scores of 54.5 points and 131.5 points, respectively (Supplementary Figure 1). When comparing the PHDI scores by gender, girls had a mean of 91.1 ± 13.5 points, while boys had a mean of 88.8 ± 12.7 points. No differences were found according to the place of residence between Barcelona and Madrid (89.2 ± 13.1 points and 91.2 ± 13.1 points, respectively).

Regarding the PHDI scores, low adherence was identified for key items such as soy and soy foods (90% below 5 points), red and processed meat (74.3% below 5 points), pulses (90% below 5 points), whole grains (40.9% below 5 points), unsaturated oils (40.9% below 5 points), and peanuts and tree nuts (40.1% below 5 points). The baseline proportion of these items within the total PHDI score is shown in Supplementary Figure 2. Reasonably, there were significant increases and decreases in most reported dietary components and food items across the PHDI quartiles (Table 2; Supplementary Figure 3). When examining PHDI attitudes toward adherence in families with migrant mothers, there was little to no difference compared to other families (Supplementary Figure 4). Participants in the highest quartile consumed more dietary fiber, potassium, and seafoods, and consumed almost double amounts of legumes, nuts, vegetables (dark & green/red & orange), and fruits compared to the lowest quartile. Increased consumption of whole grains, olive oil, and plant milks was observed in the highest quartile, while dairy, animal proteins (beef, pork, and poultry), potatoes, added sugars, and sugar-sweetened beverages were less consumed compared to the lowest quartile of the PHDI. As dietary patterns shifted toward higher PHDI adherence, participants consumed fewer calories from proteins (plant-based and animal-based proteins), along with reduced sodium and increased potassium intake. Differences in calcium, zinc, iron, and vitamin B12 intake were identified, with levels becoming more pronounced by gender, according to the PHDI (Supplementary Figure 5).

The mean follow-up of participants was 3.5 years. In the fully adjusted Cox regression models (Table 3), significant linear trends were observed, indicating a reduced risk of high BP, elevated plasma glucose, TG, total cholesterol, and non-HDL-C with increased PHDI adherence. A significant inverse association with high adherence to PHDI, whether evaluated in quartiles, and for each 20-point increase (Table 3) was observed for high BP, and increased glucose, TG, total cholesterol, and non-HDL-C. In contrast, no significant reductions for LDL-C, obesity (HR: 0.73 [95% CI: 0.48, 1.11]; p-value = 0.14), and HDL-C (HR: 1.86 [95% CI: 0.42, 8.27]; p-value = 0.42) were found by 20-point increase. When comparing the higher vs. lower PHDI adherence _{(Q4 vs Q1)}, the risk of high BP was significantly reduced by 81% (HR: 0.19 [95% CI: 0.11, 0.34]), plasma glucose by 47% (HR: 0.53 [95% CI: 0.48, 0.58]), TG by 66% (HR: 0.34 [95% CI: 0.18, 0.65]), total cholesterol by 51% (HR: 0.49 [95% CI: 0.34, 0.69]), and non-HDL-C by 74% (HR: 0.26 [95% CI: 0.13, 0.50]). Comparisons between the standardized diet and the reported diet showed no variations in the estimators (Supplementary Table 1). Limitations to study obesity and HDL-C > 40 mg/dL were attributable to the low incidence, 27 cases of obesity that made up approximately 3.3%; while out of the 35 participants who had HDL-C below 40 at baseline, 20 experienced an increase in their HDL-C concentrations. PHDI_{(Q4 vs Q1)} gender stratified analyses for elevated cardiometabolic parameters achieved statistical significance in both genders, mainly in girls (Figure 2). The results of the RCS Cox regression showed a non-significant J-shaped association (p-value for non-linearity) between the PHDI and outcomes of interest (Supplementary Figure 6).

Results of the linear mixed models assessing changes in cardiometabolic risk biomarkers and the PHDI are shown in Table 4. The PHDI _{(Q4 vs Q1)} was inversely associated with reductions of glucose (−5.23 mg/dL [95% CI: −10.35, −0.10]), TG (−2.48 mg/dL [95% CI: −3.65, −1.30]), and BMI z-score (−0.02 [95% CI: −0.03, 0.00]). In addition, we found statistically significant inverse associations between PHDI _{(Q3 vs Q1)}, SBP (−3.83 mmHg [95% CI: −3.98, −3.67]), and DBP (−3.83 mmHg [95% CI: −3.98, −3.67]) in participants with high BP at baseline. Overall, trends were observed in the other cardiometabolic risk parameters; however, no association was observed when comparing them as continuous variables.

Our results indicate that a high PHDI adherence_{(Q4 vs Q1)} is associated with a risk reduction of high BP, increased plasma glucose, TG, total cholesterol, and non-HDL-C by 81, 47, 66, 51, and 74% respectively; and that high adherence to this dietary pattern is longitudinally associated with reductions of glucose (−5.23 mg/dL), TG (−2.48 mg/dL), and BMI z-score (−0.02) in adolescents. In contrast, when evaluating the PHDI _{(Q4 vs Q1)} and its relationship with LDL-C, total cholesterol, WHtR, SBP, and DBP as continuous variables in linear mixed models, no significant association is found. However, trends suggest a potential inverse association, which may reflect a healthier nutritional status. The analyses are robust, with most associations remaining significant after adjustments for various confounders and after sensitivity analyses. We found that when estimating the nutritional requirements (28) for energy, protein, zinc, iron, and vitamin B12 among adolescents with higher adherence to the PHDI, their intake was well within the recommended range. However, calcium intake was insufficient in female adolescents (4.4% below the requirements) among the highest quartiles. Promotion of calcium-rich plant foods and intake of mineral water as part of a healthy, balanced diet may facilitate reaching an adequate concentration of calcium.

Evidence that adherence to the PHDI is associated with changes in lipid profile, glucose, BP, or anthropometric parameters among adolescents is limited. In fact, consistent with our findings, a high PHDI adherence _{(10-point increase)} was associated with lower odds of hypertension (OR: 0.87 [95% CI: 0.79, 0.96]) (29), lower odds of increased total cholesterol (OR: 0.88 [95% CI: 0.78, 0.99]), and a higher Ideal Cardiovascular Health score. One study also found reductions in anthropometric parameters with PHDI adherence (30). Adjusted mean estimates revealed inverse associations between PHDI adherence and body weight (0.98 kg [95% CI: 0.97, 0.99]), BMI (0.99 kg/m² [95% CI: 0.97, 0.99]), fat-free mass index (0.99 [95% CI: 0.99, 0.99]), waist circumference (0.99 cm [95% CI: 0.98, 0.99]), and body fat (0.98% [95% CI: 0.96, 0.99]) in an European cohort (30). Our results, regarding the excessive consumption of red and processed meats (Supplementary Figure 3) and adherence to the PHDI, are consistent with findings from studies in other young populations (31, 32). A surprising finding was that attitudes toward PHDI adherence were similar, including in families with a migrant background, as shown in Supplementary Figure 4. Some factors, such as socioeconomic level, duration of residence, and the role of country of origin (Mediterranean or non-Mediterranean), may promote dietary acculturation by adopting a healthy or a detrimental “Westernized” dietary pattern (33, 34).

The studies summarized below examine the effects of plant-based diets, such as the Mediterranean diet and the Dietary Approaches to Stop Hypertension (DASH) diet, on cardiometabolic biomarkers in pediatric populations, which underscore the relevance of this topic (35–40). These diets include food items similarly recommended in the Planetary Health Diet, but differing in the emphasis on sustainable eating. Results of a systematic review and meta-analysis evaluating the mean differences following the Mediterranean diet showed a significant inverse effect on SBP (−4.75 mm Hg [95% CI: −8.97,−0.52]), TG (−16.42 mg/dL [95% CI: −27.57, −5.27]), total cholesterol (−9.06 mg/dL [95% CI: −15.65, −2.48]), and LDL-C (−10.48 mg/dL [95% CI: −17.77, −3.19]), while increasing HDL-C (2.24 mg/dL [95% CI: 0.34, 4.14]) (41). Plant-rich diets have also shown to significantly reduce the odds of developing hypertension (OR, 0.63 [95% CI: 0.41, 0.97]) (35), mitigating the harmful effects of oxidative stress and inflammation, which are exacerbated by prolonged high BP (36). Significant reductions in SBP have been reported in two interventional studies: a decrease of 2.7 mmHg (p-value = 0.03) (36) and 10.4 mmHg (p-value <0.01) (37), both of them following the DASH diet. The Framingham Children’s study similarly observed reductions in mean SBP and DBP (4.60 mmHg and 1.12 mmHg, respectively) with a diet rich in fruits and vegetables (>4 servings/day) and low in dairy products (< 2 servings/day) (38). Other plant-based diets, such as the Mediterranean diet, have also shown significant improvements in lipid profile after a lifestyle intervention (39, 40). For example, reductions in total cholesterol (23.5 mg/dL), LDL-C (21.5 mg/dL), and non-HDL-C (21.5 mg/dL) were observed in children diagnosed with primary hypercholesterolemia (39), and in total cholesterol (25.5 mg/dL), LDL-C (22.0 mg/dL), and TG (12.0 mg/dL) (40).

Several reasons could explain the beneficial results in cardiometabolic biomarkers following this healthy pattern. One is the influence of low added sugars consumption (≤25 g/day of added sugars) (42), along with high consumption of whole grains, legumes, fruits, and vegetables, which has been shown to mitigate postprandial glucose excursions (41). A second reason underlies the consumption of phytonutrients, which have shown to activate β-oxidation, regulate satiety, and modulate energy intake (5, 43). These compounds can also induce thermogenesis in brown adipose tissue, and mobilize stored fat (5), while exhibiting antioxidant, anti-inflammatory, and immune-modulating properties (43, 44). A third reason involves the consumption of dietary fibers, such as β-glucans, arabinoxylans, and lignins, that play a crucial role in cardiometabolic health. They reduce the absorption of lipids and carbohydrates (45), control circulating LDL-C by inhibiting bile acid reabsorption (45), increase bacterial diversity (44), enhance intestinal barrier integrity (45), and promote the growth and metabolism of beneficial commensal Clostridia (Firmicutes) (46). High fermentation by major butyrate-producing bacteria, such as Faecalibacterium, further influences lipid metabolism through G-protein coupled receptors 41 and 43 (45, 47).

The strengths of the present study include its large-scale and long-term assessment of the PHDI and subsequent cardiometabolic risk screening during adolescence. Our results also addressed the controversies regarding nutrient intake, of particular concern in plant-based diets, in this population. Handling missing values in studied variables helped retain precise information for analysis, even after data loss occurred throughout the follow-up. Capillary blood measurements enabled us to efficiently assess cardiometabolic blood biomarkers, while wearable accelerometers offered objective measurements of physical activity and sleep. Repeated dietary assessments, along with the use of time-varying covariates to capture changes in dietary habits and health status, helped to reduce intraindividual variation over time. Furthermore, the validated FFQ used in this study showed good reproducibility and validity, and participants with extreme energy intakes were excluded to optimize the reliability of dietary information. By employing various data-driven analysis techniques, we were able to identify meaningful patterns and relationships.

However, the study also has some limitations. First, although BP readings were taken during the visits following standardized protocols, hypertension diagnoses were not confirmed through clinical records. Second, the use of proxy reporters for diet and other covariates may induce potential residual confounding in the analysis. The results may not be generalizable to other populations because participants were from a Mediterranean country. Caution should be exercised when interpreting glucose results, as some adolescents may have been assessed in a non-fasting state despite reporting that they were fasting. Additionally, the study design allows for the observation of the associations between PHDI adherence and outcomes, but it does not establish causality. It reflects the real-world conditions of the population rather than controlled experimental scenarios. Finally, the association between PHDI adherence, cardiometabolic health, and high BP was not a predefined endpoint of the SI! Program. This study makes the findings exploratory, requiring further research regarding PHDI, adolescent health, and educational strategies.

We observed strong inverse associations between higher PHDI adherence and the incidence of high BP, as well as several cardiometabolic risk factors. These findings suggest that a healthy plant-based diet, rich in phytochemicals and dietary fiber, promotes cardiometabolic health from an early age. This approach should mitigate the progression of cardiovascular diseases in younger populations while also reducing the environmental impact. Practical implications of these results include incorporating PHDI recommendations into school meals and education campaigns targeting adolescents and their families to promote the dual benefits of this diet for health and planetary sustainability.