Polyphenol-rich extra-virgin olive oil influences bioenergetics and may modulate NLRP3 inflammasome in PBMCs from children and adolescents with obesity Authors: Maria Laura Matrella, Luigi Malaspina, Mariangela Chiarito, Francesco Limongelli, Rossana Franzin, Laura Piacente, Daniela De Benedictis, Roberta Tardugno, Tiziana Cocco, Maria Lisa Clodoveo, Paola Pontrelli, Anna Signorile, Filomena Corbo, Maria Felicia Faienza Published: 2026 ABSTRACT Paediatric obesity is characterized by chronic inflammation and immunometabolic dysregulation. Extra-virgin olive oil (EVOO) polyphenols exert antioxidant and immunomodulatory effects that may counteract obesity-related alterations. A 12-week controlled nutritional intervention was conducted in 84 children and adolescents with obesity (BMI > 95th percentile), comparing a Mediterranean diet supplemented with either high-polyphenol EVOO (P-EVOO) or refined olive oil (R-OO). Anthropometric and biochemical parameters were evaluated at baseline and study completion. In peripheral blood mononuclear cell (PBMC) of a sub-cohort, bioenergetic analyses, gene expression related to mitochondrial biogenesis, antioxidant defence, and NLRP3 gene expression were assessed. Both groups exhibited significant reductions in BMI, waist circumference, LDL cholesterol, triglycerides and uric acid. A significant reduction in insulin resistance was observed exclusively in the P-EVOO group. In PBMCs, P-EVOO enhanced mitochondrial respiration and ATP production, associated with a metabolic shift toward oxidative phosphorylation. This was accompanied by increased mitochondrial biogenesis and activation of the Nrf2–SOD2 antioxidant axis. P-EVOO upregulated NLRP3, IL-1β, CASP8 and CASP1 gene expression, without IL-18 or GSDMD increase, suggesting a distinct expression pattern of alternative NLRP3 inflammasome-associated markers. These findings indicate that P-EVOO could ameliorate metabolic health and mitochondrial bioenergetics in paediatric obesity, promoting an oxidative immunometabolic phenotype potentially associated with controlled inflammasome transcriptional regulation.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-51300-9. Introduction Childhood obesity represents a major global public health concern, affecting millions of children worldwide, and is recognized as a chronic, multifactorial disease driven by the complex interaction of biological, behavioral, environmental, social, and economic determinants1. It frequently tracks into adulthood, and the adverse consequences of excessive adiposity can manifest early through the development of chronic conditions, including cardiovascular disease (CVD), type 2 diabetes (T2D), non-alcoholic fatty liver disease (NAFLD), and a range of psychological disorders2–4. Obesity is recognized as a chronic inflammatory condition, evidenced by increased immune cell infiltration into adipose tissue, and the consequently elevated release of circulating proinflammatory cytokines. In normal-weight individuals, adipose immune cells maintain tissue balance by regulating inflammation, whereas in chronic obesity they create a persistent pro-inflammatory state that impairs adipocyte insulin sensitivity and drives systemic insulin resistance5. This process is underpinned by metabolic reprogramming, in which immune cells shift from mitochondrial oxidative metabolism in the resting state to glycolysis during inflammatory activation6. However, conflicting results are reported for mitochondrial function of immune cells in obesity. Increased mitochondrial oxidative phosphorylation (OXPHOS) capacity has been observed in peripheral blood mononuclear cells (PBMCs) of obese young adults, correlating with insulin resistance, systemic inflammation, and hepatic steatosis7. Similarly, in childhood obesity, alterations in the bioenergetic profile and mitochondrial dysfunction have been reported across different immune cell subsets. For instance, in PBMCs from obese children, the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) ratio positively correlates with body mass index (BMI) and negatively with glycosylated hemoglobin (HbA1c)8. Moreover, PBMCs from overweight or obese insulin-sensitive children display a metabolic shift toward glycolysis compared with their normal-weight insulin-sensitive peers9. The NLRP3 inflammasome, comprising the NLRP3 sensor, the adaptor ASC, and the effector protease caspase-1, represents a key molecular link connecting obesity-related adipose inflammation with the development of insulin resistance9. This cytosolic complex detects a broad range of metabolic danger mediators, both endogenous and exogenous, including free fatty acids, reactive oxygen species (ROS), and extracellular ATP. Through this stress-sensing capacity, NLRP3 orchestrates inflammatory responses that contribute to obesity-associated metabolic dysfunction. Experimental evidence indicates that long-chain omega-3 polyunsaturated fatty acids derived from fish oil attenuate its activation10. Among the various dietary patterns that have demonstrated beneficial effects on human health, the Mediterranean diet (MD) has proven particularly effective in the management of several conditions, including obesity11,12. The health-promoting effects of the MD are largely attributable to extra-virgin olive oil (EVOO). EVOO is considered a functional food showing beneficial effects that could be ascribed to a synergic effect of monounsaturated fatty acids, particularly oleic acid, and phenolic compounds. In particular, phenolic compounds contained in EVOO13 showed ability to reduce inflammation milieu of PBMCs from obese children14. Recent studies have re-evaluated the role of its minor fraction, mainly composed of polyphenols with anti-inflammatory15 and antioxidant properties13, which appears to act synergistically with the major fraction represented by monounsaturated fatty acids, particularly oleic acid16. In addition, polyphenols contained in EVOO can modulate mitochondrial function17. Considering this evidence, this longitudinal study aimed to evaluate the effects of MD supplemented with polyphenol-rich extra-virgin olive oil (P-EVOO), compared with MD supplemented with refined olive oil (R-OO) obtained through standard industrial refining processes that remove the polyphenolic fraction, without changing the fatty acid composition, on anthropometric and biochemical parameters in a cohort of children and adolescents with obesity. In addition, in a sub-cohort of participants, we investigated the effects of both dietary interventions on PBMC bioenergetics and on the expression of genes involved in mitochondrial biogenesis, antioxidant defence, and NLRP3 inflammasome-related genes transcriptional regulation. Results Biochemical and anthropometric parameters in the enrolled cohort Of the 84 subjects recruited (40 males), 59 completed the study, and one participant was excluded following a new diagnosis of diabetes. The overall dropout rate was 29.8%. At the beginning of the study (T0), 24 subjects (13 males) were assigned to the R-OO group, and 34 subjects (14 males) to the P-EVOO group. Table 1 reports the characteristics of the study population at T0 and after 12-week (T-END) controlled nutritional intervention with MD supplemented with high-polyphenol EVOO (P-EVOO) refined olive oil or (R-OO). At baseline, the two groups were homogeneous with respect to anthropometric and biochemical parameters. Table 1Characteristics of R-OO and P-EVOO group at T0 and T-END.T0T- END P Age (yr)Group R-OO11.3 ± 3.111.6 ± 3.1–Group P-EVOO10.7 ± 2.911.0 ± 2.9–Height (SDS)Group R-OO0.59 (-0.5-0.8)0.5 (-0.5-1.04)0.26bGroup P-EVOO0.4 (-0.2-1.1)0.5 (-0.2-1.1)0.544bWeight (SDS)Group R-OO2.2 ± 0.71.98 ± 0.89< 0.001aGroup P-EVOO2.03 ± 0.921.84 ± 1.02< 0.001aBMI (SDS)Group R-OO2.4 ± 0.62.2 ± 0.8< 0.001aGroup P-EVOO2.16 ± 0.62 ± 0.8< 0.001aWC (cm)Group R-OO96.5 ± 15.793.7 ± 14.9< 0.001aGroup P-EVOO92.3 ± 14.189.1 ± 14.6< 0.001aAST (u/L)Group R-OO24 (20-28.8)21 (16–24)0.017bGroup P-EVOO22.5 (18-26.25)21 (17-23.25)0.147bALT (u/L)Group R-OO19 (16-29.3)18 (13 − 2)0.111bGroup P-EVOO21 (17-26.8)17.5 (14-27.8)0.237bγ-GT (u/L)Group R-OO18.5 (13.3–21)15.5 (11.8–18.3)0.018bGroup P-EVOO19.5 (15.3–22.8)16 (13-20.5)0.139bTC (mg/dl)Group R-OO146.5 (113.5-168.5)141 (117–158)0.353bGroup P-EVOO146 (135–166)135 (130-153.3)0.085bHDL (mg/dl)Group R-OO41.7 (34.7–51.5)42 (36.7–49.6)0.681bGroup P-EVOO44.5 (40.7–54.2)43.2 (37.4–51.4)0.782bLDL-C (mg/dl)Group R-OO100.5 ± 3593.7 ± 27.3< 0.001aGroup P-EVOO98.9 ± 28.794.1 ± 20.4< 0.001aTG (mg/dl)Group R-OO88.5 (74.8-132.8)86 (53–102)0.009bGroup P-EVOO82 (63.5–99)76 (57.5–108)0.722bTG/HDLGroup R-OO2.12 (1.6-3)1.8 (1.2–2.5)0.018bGroup P-EVOO1.7 (1.1–2.6)1.7 (1.3–2.4)0.622bGlucose (mg/dl)Group R-OO86.9 ± 4.787.8 ± 7.80.557aGroup P-EVOO84.3 ± 6.484.4 ± 60.883aInsulinGroup R-OO18.4 (12.9–23.0)13.2 (10.1–23.8)0.772bGroup P-EVOO17.9 (11.1–22.4)15.4 (11.4–20.9)0.115bHbA1c (mmol/l)Group R-OO33.5 (30.8–35.5)32 (29.8–35)0.234bGroup P-EVOO33 (30–36)32 (29–33)0.476bHOMA-IRGroup R-OO4.0 (5.9-5.0)2.3 (1.9–4.7)0.761bGroup P-EVOO3.7 (2.3–5.1)2.9 (1.6–4.2)0.029bUric acid (mg/dl)Group R-OO5.04 ± 1.25.04 ± 1.5< 0.001aGroup P-EVOO4.9 ± 1.14.8 ± 0.95< 0.001aIL-1βGroup R-OO12.7 (1.95–34.9)10.9 (3-30.7)0.510bGroup P-EVOO9.6 (2.6–16.1)6.9 (0.4–20.4)0.904bIL-6Group R-OO4 (2-8.2)4.6 (2.1–23.6)0.678bGroup P-EVOO5.5 (1.8–9.2)2.6 (1.5–7.8)0.917bCRPGroup R-OO3.8 ± 0.54.6 ± 1.90.003aGroup P-EVOO5.8 ± 4.77.9 ± 9.50.017aKIDMEDGroup R-OO7 (4.5-9)7.5 (4–11)0.012bGroup P-EVOO6 (4–7)6.5 (5–8)0.02bacomparison performed by Student’s t-test.bcomparison performed by Mann–Whitney test.WC, waist circumference; AST, aspartate transaminase; ALT, alanine transaminase; γ-GT, γ-glutamil transferase; TC, total cholesterol; HDL, high density lipoprotein; LDL, low density lipoprotein; TG, triglycerides; IL-1β, interleukin 1-β; IL-6, interleukin 6; BMI, body mass index; CRP, C reactive protein; Table 1Characteristics of R-OO and P-EVOO group at T0 and T-END.T0T- END P Age (yr)Group R-OO11.3 ± 3.111.6 ± 3.1–Group P-EVOO10.7 ± 2.911.0 ± 2.9–Height (SDS)Group R-OO0.59 (-0.5-0.8)0.5 (-0.5-1.04)0.26bGroup P-EVOO0.4 (-0.2-1.1)0.5 (-0.2-1.1)0.544bWeight (SDS)Group R-OO2.2 ± 0.71.98 ± 0.89< 0.001aGroup P-EVOO2.03 ± 0.921.84 ± 1.02< 0.001aBMI (SDS)Group R-OO2.4 ± 0.62.2 ± 0.8< 0.001aGroup P-EVOO2.16 ± 0.62 ± 0.8< 0.001aWC (cm)Group R-OO96.5 ± 15.793.7 ± 14.9< 0.001aGroup P-EVOO92.3 ± 14.189.1 ± 14.6< 0.001aAST (u/L)Group R-OO24 (20-28.8)21 (16–24)0.017bGroup P-EVOO22.5 (18-26.25)21 (17-23.25)0.147bALT (u/L)Group R-OO19 (16-29.3)18 (13 − 2)0.111bGroup P-EVOO21 (17-26.8)17.5 (14-27.8)0.237bγ-GT (u/L)Group R-OO18.5 (13.3–21)15.5 (11.8–18.3)0.018bGroup P-EVOO19.5 (15.3–22.8)16 (13-20.5)0.139bTC (mg/dl)Group R-OO146.5 (113.5-168.5)141 (117–158)0.353bGroup P-EVOO146 (135–166)135 (130-153.3)0.085bHDL (mg/dl)Group R-OO41.7 (34.7–51.5)42 (36.7–49.6)0.681bGroup P-EVOO44.5 (40.7–54.2)43.2 (37.4–51.4)0.782bLDL-C (mg/dl)Group R-OO100.5 ± 3593.7 ± 27.3< 0.001aGroup P-EVOO98.9 ± 28.794.1 ± 20.4< 0.001aTG (mg/dl)Group R-OO88.5 (74.8-132.8)86 (53–102)0.009bGroup P-EVOO82 (63.5–99)76 (57.5–108)0.722bTG/HDLGroup R-OO2.12 (1.6-3)1.8 (1.2–2.5)0.018bGroup P-EVOO1.7 (1.1–2.6)1.7 (1.3–2.4)0.622bGlucose (mg/dl)Group R-OO86.9 ± 4.787.8 ± 7.80.557aGroup P-EVOO84.3 ± 6.484.4 ± 60.883aInsulinGroup R-OO18.4 (12.9–23.0)13.2 (10.1–23.8)0.772bGroup P-EVOO17.9 (11.1–22.4)15.4 (11.4–20.9)0.115bHbA1c (mmol/l)Group R-OO33.5 (30.8–35.5)32 (29.8–35)0.234bGroup P-EVOO33 (30–36)32 (29–33)0.476bHOMA-IRGroup R-OO4.0 (5.9-5.0)2.3 (1.9–4.7)0.761bGroup P-EVOO3.7 (2.3–5.1)2.9 (1.6–4.2)0.029bUric acid (mg/dl)Group R-OO5.04 ± 1.25.04 ± 1.5< 0.001aGroup P-EVOO4.9 ± 1.14.8 ± 0.95< 0.001aIL-1βGroup R-OO12.7 (1.95–34.9)10.9 (3-30.7)0.510bGroup P-EVOO9.6 (2.6–16.1)6.9 (0.4–20.4)0.904bIL-6Group R-OO4 (2-8.2)4.6 (2.1–23.6)0.678bGroup P-EVOO5.5 (1.8–9.2)2.6 (1.5–7.8)0.917bCRPGroup R-OO3.8 ± 0.54.6 ± 1.90.003aGroup P-EVOO5.8 ± 4.77.9 ± 9.50.017aKIDMEDGroup R-OO7 (4.5-9)7.5 (4–11)0.012bGroup P-EVOO6 (4–7)6.5 (5–8)0.02bacomparison performed by Student’s t-test.bcomparison performed by Mann–Whitney test.WC, waist circumference; AST, aspartate transaminase; ALT, alanine transaminase; γ-GT, γ-glutamil transferase; TC, total cholesterol; HDL, high density lipoprotein; LDL, low density lipoprotein; TG, triglycerides; IL-1β, interleukin 1-β; IL-6, interleukin 6; BMI, body mass index; CRP, C reactive protein; Characteristics of R-OO and P-EVOO group at T0 and T-END. acomparison performed by Student’s t-test. bcomparison performed by Mann–Whitney test. WC, waist circumference; AST, aspartate transaminase; ALT, alanine transaminase; γ-GT, γ-glutamil transferase; TC, total cholesterol; HDL, high density lipoprotein; LDL, low density lipoprotein; TG, triglycerides; IL-1β, interleukin 1-β; IL-6, interleukin 6; BMI, body mass index; CRP, C reactive protein; To assess the effects of the different nutritional interventions, all parameters were analyzed within each group by comparing T-END with baseline (T0). At T-END, the R-OO group showed a significant reduction in weight, BMI-SDS, WC, LDL-C, triglycerides, and urate. The P-EVOO group demonstrated a statistically significant decrease in weight, BMI, WC, LDL-C, urate, and HOMA-IR. Both groups exhibited a significant improvement in KIDMED scores. Effects of P-EVOO on the bioenergetics profile in PBMCs of obese children To investigate the effects of different oil supplementations on the bioenergetic profile of PBMCs, a sub-cohort of 20 participants was randomly selected from the study population. A systematic analysis was performed using Seahorse technology to assess the mitochondrial OXPHOS activity, measured as oxygen consumption rate (OCR), and the extracellular acidification rate (ECAR), primarily reflecting the conversion of pyruvate to lactate, essentially following the manufacturer’s Mito Stress protocol. Analysis of OCR and ECAR in PBMCs at baseline (T0) showed no differences between the R-OO and P-EVOO groups (Supplementary Fig. A), confirming the homogeneity of the two groups at T0. A two-way repeated measures ANOVA of basal respiration, which reflects the minimal oxygen consumption required to sustain essential cellular functions, showed no significant change in PBMCs from the R-OO and P-EVOO groups (Fig. 1, Supplementary Table A1). Coupled respiration, which represents oxygen utilization directly linked to ATP synthesis, was not affected by time, time x group intervention and group intervention (Supplementary Table A1). However, post-hoc analysis using Sidak’s showed that coupled respiration increased in the P-EVOO from T0 to T-END (p = 0.0275) (Fig. 1, Supplementary Table A1). In contrast, PBMCs from the R-OO group showed no significant differences in either basal or coupled respiration (Fig. 1, Supplementary Table A1). These findings suggest an increased reliance on OXPHOS to meet their energetic demands in PBMCs from P-EVOO group. Fig. 1Increased OXPHOS capacity by P-EVOO supplemented MD . PBMCs from obese subjects in basal respiration, coupled respiration, maximal respiration, spare respiratory capacity and proton leak in PBMCs from pediatric obese patients were assessed using the Seahorse XF24 analyzer and normalized on protein content. The bioenergetic parameters were measured before (T0) and after 12 weeks (T-END) of MD supplemented with R-OO (group R-OO) (n = 9) or P-EVOO (group P-EVOO) (n = 11). Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’s multiple comparison test (ns = not significant; ***p < 0.005, *p < 0.05). Increased OXPHOS capacity by P-EVOO supplemented MD . PBMCs from obese subjects in basal respiration, coupled respiration, maximal respiration, spare respiratory capacity and proton leak in PBMCs from pediatric obese patients were assessed using the Seahorse XF24 analyzer and normalized on protein content. The bioenergetic parameters were measured before (T0) and after 12 weeks (T-END) of MD supplemented with R-OO (group R-OO) (n = 9) or P-EVOO (group P-EVOO) (n = 11). Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’s multiple comparison test (ns = not significant; ***p < 0.005, *p < 0.05). Maximal respiration, recorded after addition of the uncoupler FCCP, which increases OCR to its maximal rate, resulted affected by time (p = 0.0007) with Šídák post-hoc analysis confirming a significant increase in P-EVOO group (p = 0.0003) (Supplementary Table A1). At T-END significative differences were observed in maximal respiration between the two groups (p = 0.0369) (Fig. 1, Supplementary Table A1). Accordingly, the spare capacity, defined as the difference between FCCP-stimulated OCR and basal OCR, was significantly affected by time (p = 0.0003) and group intervention (p = 0.0470) (Supplementary Table A1 and a significant increase in P-EVOO group (p = 0.0002) was confirmed by the Šídák post-hoc analysis a. At T-END significative differences were observed in spare capacity between groups (p = 0.0298) (Fig. 1, Supplementary Table A1). Overall, these findings are consistent with a bioenergetic profile of PBMC from P-EVOO group predominantly sustained by OXPHOS, together with a marked increase in mitochondrial reserve capacity, reflecting enhanced adaptability to increased energy demands and stress conditions. These data also allowed to evaluate the proton leak which did not show any significant difference (Fig. 1). The Mito Stress Test also provides information on ECAR, which predominantly reflects glycolytic flux. ECAR values, corrected for residual activity in the presence of the glycolysis inhibitor 2-DG, are reported in Fig. 2. PBMCs from both the R-OO and P-EVOO groups showed no significant differences in basal glycolysis (defined as the difference between basal ECAR and 2-DG–inhibited ECAR) (Fig. 2). However, glycolytic max capacity, calculated as the difference between FCCP-stimulated ECAR and 2-DG–inhibited ECAR, as well as, glycolytic reserve capacity, defined as the difference between FCCP-stimulated and basal ECAR, resulted affected by time (p = 0.0201 and p = 0.0053 respectively) (Supplementary Table A1. Šídák post-hoc analysis showed a significant increase of glycolytic max capacity and glycolytic reserve capacity in PBMC from P-EVOO group at T-END respect to T0 (Fig. 2, Supplementary Table A1). No changes in glycolytic capacity or glycolytic reserve were observed at T-END in PBMCs from the R-OO group (Fig. 2, Supplementary Table A1), while significant differences were observed in glycolytic reserve capacity at T-END between R-OO and P-EVOO group (p = 0.0348) (Fig. 2, Supplementary Table A1). Fig. 2Increased glycolytic capacity by P-EVOO supplemented MD in PBMCs of obese patients. Glycolysis, glycolytic max capacity and glycolytic reserve capacity were assessed using Seahorse XF24 analyzer and normalized on protein content in PBMCs from paediatric obese patients, before (T0) and after 12 weeks (T-END) of MD supplemented with R-OO (group R-OO) (n = 19) or P-EVOO (group P-EVOO) (n = 11). Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’s multiple comparison test (ns = not significant; **p < 0.01, *p < 0.05). Increased glycolytic capacity by P-EVOO supplemented MD in PBMCs of obese patients. Glycolysis, glycolytic max capacity and glycolytic reserve capacity were assessed using Seahorse XF24 analyzer and normalized on protein content in PBMCs from paediatric obese patients, before (T0) and after 12 weeks (T-END) of MD supplemented with R-OO (group R-OO) (n = 19) or P-EVOO (group P-EVOO) (n = 11). Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’s multiple comparison test (ns = not significant; **p < 0.01, *p < 0.05). Interesting, analysis of the bioenergetic profiles derived from maximal cellular oxygen consumption (OCR) and glycolytic max capacity (ECAR) revealed that OCR/ECAR was affected by time (p = 0.0313) (Supplementary Table A1). Furthermore, Šídák post-hoc analysis showed that PBMCs from the P-EVOO group exhibited an increased OCR/ECAR ratio at T-END (p = 0.0389). No changes of this ratio were observed at T-END in the R-OO group (Fig. 3, Supplementary Table A1). Fig. 3Shift toward oxidative metabolism by P-EVOO supplemented MD in PBMCs from obese subjects. Rate of maximal cellular oxygen consumption (OCR) and glycolytic max capacity (ECAR). Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’s multiple comparison test (ns = not significant; *p < 0.05). Shift toward oxidative metabolism by P-EVOO supplemented MD in PBMCs from obese subjects. Rate of maximal cellular oxygen consumption (OCR) and glycolytic max capacity (ECAR). Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’s multiple comparison test (ns = not significant; *p < 0.05). To assess ATP production from both glycolysis and oxidative phosphorylation, we applied a validated Seahorse-based protocol that simultaneously measures oxygen consumption and extracellular acidification in the same sample18,19. This approach enables direct quantification of ATP generated through mitochondrial respiration (mitoATP) and glycolysis (glycoATP), providing a robust framework for evaluating shifts in cellular energy metabolism. As shown in Fig. 4 and Supplementary Table A2 only the P-EVOO group exhibited a significant increase in mitochondrial proton production rate (MitoPPR) at T-END compared with T0 (p = 0.0052), indicating enhanced respiratory CO₂-derived proton flux. In contrast, no significant changes were detected in the R-OO group (Fig. 4A). Glycolytic proton production rate (GlycoPPR) remained unchanged in both groups, suggesting that glycolytic activity was not markedly affected by either intervention (Fig. 4B). Consistently, two-way repeated measures ANOVA revealed that mitochondrial ATP production (MitoATP) was affected by time (p < 0.0001) (Supplementary Table A2) and increased significantly after P-EVOO intake (P < 0.0001), whereas no differences were observed in the R-OO group (Fig. 4C). Glycolytic ATP production (GlycoATP) did not vary significantly in either condition (Fig. 4D). Overall, these findings indicate that P-EVOO supplementation selectively enhances mitochondrial oxidative metabolism and ATP generation in PBMCs of paediatric obese subjects, without altering glycolytic output. Fig. 4Increased oxidative metabolism and mitochondrial ATP production by P-EVOO supplemented MD in PBMCs from paediatric obese subjects. (A,B) Proton production rates (PPR) from respiratory CO2 production (MitoPPR) (A) and glycolytic lactate (GlycoPPR) (B) in PBMCs from paediatric obese subjects, before (T0) and after 12 weeks (T-END) of MD supplemented with either R-OO (group R-OO) (n = 7) or P-EVOO (group P-EVOO) (n = 10). PPR was expressed as pmol H+/min/µg protein. All data are mean ± SEM. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’smultiple comparison test (**p < 0.01; ns = not significant). (C,D) MitoATP (C) and GlycoATP (D) data, means ± SEM, are shown as stacked bars. ATP production rate was expressed as pmol ATP/min/µg protein. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’s multiple comparison test (*p < 0.05 and ****p < 0.0001; ns = not significant). Increased oxidative metabolism and mitochondrial ATP production by P-EVOO supplemented MD in PBMCs from paediatric obese subjects. (A,B) Proton production rates (PPR) from respiratory CO2 production (MitoPPR) (A) and glycolytic lactate (GlycoPPR) (B) in PBMCs from paediatric obese subjects, before (T0) and after 12 weeks (T-END) of MD supplemented with either R-OO (group R-OO) (n = 7) or P-EVOO (group P-EVOO) (n = 10). PPR was expressed as pmol H+/min/µg protein. All data are mean ± SEM. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’smultiple comparison test (**p < 0.01; ns = not significant). (C,D) MitoATP (C) and GlycoATP (D) data, means ± SEM, are shown as stacked bars. ATP production rate was expressed as pmol ATP/min/µg protein. Statistical analysis was performed using a mixed-effects model (REML), followed by Sidak’s multiple comparison test (*p < 0.05 and ****p < 0.0001; ns = not significant). Effects of P-EVOO on mitochondrial biogenesis and cellular antioxidant response To determine whether the functional improvement of mitochondrial oxidative metabolism was accompanied by coordinated molecular adaptations within the mitochondrial network, we next evaluate the effects of P-EVOO on mitochondrial biogenesis in PBMCs from obese subjects, we measured the mRNA expression levels of key regulatory genes—PPARGC1A, TFAM, NRF1, MT-ATP6, and COXIV—before (T0) and after three months (T-END) of oil intake. Two-way repeated measures ANOVA revealed a significant effect of time x group intervention (p < 0.0001) and group intervention (p < 0.0001), of all genes involved in mitochondrial biogenesis (Supplementary Table A3). Post-hoc analysis using Šídák’s multiple comparisons test showed that R-OO consumption led to a significant downregulation of PPARGC1A (p < 0.0001) and NRF1 (p = 0.0117) at T-END compared with T0, whereas the expression of TFAM, COXIV, and MT-ATP6 remained unchanged. In contrast, PBMCs from the P-EVOO group displayed a significant increase of PPARGC1A (p < 0.0001), MT-ATP6 (p = 0.0022), and COXIV (p = 0.0178) at T-END relative to T0 (Fig. 5, Supplementary Table A3). At T-END significative differences were observed between groups for all mitochondrial biogenesis regulatory genes (Fig. 5, Supplementary Table A3). Fig. 5Increased RNA levels of key regulators of mitochondrial biogenesis by P-EVOO supplemented MD in PBMCs from paediatric obese subjects. Relative mRNA levels of PPARGC1A (PGC-1α), TFAM, NRF1, MTATP6 and COXIV were determined in PBMCs from paediatric obese subjects, before (T0) and after 12 weeks (T-END) of MD supplemented with either R-OO (group R-OO) (n = 11) or P-EVOO (group P-EVOO) (n = 18). mRNA levels were quantified by RT-qPCR from total RNA normalized to GAPDH. Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using two-way repeated measures ANOVA followed by Sidak’s multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001; ns = not significant). Increased RNA levels of key regulators of mitochondrial biogenesis by P-EVOO supplemented MD in PBMCs from paediatric obese subjects. Relative mRNA levels of PPARGC1A (PGC-1α), TFAM, NRF1, MTATP6 and COXIV were determined in PBMCs from paediatric obese subjects, before (T0) and after 12 weeks (T-END) of MD supplemented with either R-OO (group R-OO) (n = 11) or P-EVOO (group P-EVOO) (n = 18). mRNA levels were quantified by RT-qPCR from total RNA normalized to GAPDH. Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using two-way repeated measures ANOVA followed by Sidak’s multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001; ns = not significant). Given the established association between obesity and oxidative stress20 we investigated the effects of P-EVOO on the expression of antioxidant response-related genes including NFE2L2 (NRF2), SOD2, and GPX1 in PBMCs isolated from obese subjects. As shown in Fig. 6, antioxidant response related genes were significantly affected by time, group and their interaction (Supplementary Table A4) and Šídák’s analysis showed that R-OO intervention is associated with a significant reduction in GPX1 expression (p = 0.0080), while no significant changes in NFE2L2, a transcription factor that regulates the expression of antioxidant proteins, or SOD2, a mitochondrial superoxide dismutase, were observed (Supplementary Table A4). In contrast, P-EVOO resulted associated with a promotion of NFE2L2 (p < 0.0001) and SOD2 (p = 0.0002) mRNA levels at T-END compared with T0. At T-END significative differences were observed between groups for NFE2L2 and SOD2 mRNA levels (Fig. 6, Supplementary Table A4). Fig. 6Increased RNA levels of antioxidant enzymes by P-EVOO supplemented MD in PBMCs from paediatric obese subjects. Relative mRNA levels of NRF2, SOD2 and GPX1 were determined by RT-qPCR in PBMCs from paediatric obese subjects, before (T0) and after 12 weeks (T-END) of MD supplemented with either R-OO (group R-OO) (n = 11) or P-EVOO (group P-EVOO) (n = 18). mRNA levels were quantified by RT-qPCR from total RNA, normalized to GAPDH. Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using two-way repeated measures ANOVA followed by Sidak’s multiple comparison test (**p < 0.01, ***p < 0.001 and ****p < 0.0001; ns = not significant). Increased RNA levels of antioxidant enzymes by P-EVOO supplemented MD in PBMCs from paediatric obese subjects. Relative mRNA levels of NRF2, SOD2 and GPX1 were determined by RT-qPCR in PBMCs from paediatric obese subjects, before (T0) and after 12 weeks (T-END) of MD supplemented with either R-OO (group R-OO) (n = 11) or P-EVOO (group P-EVOO) (n = 18). mRNA levels were quantified by RT-qPCR from total RNA, normalized to GAPDH. Box and whisker plots show the median (horizontal line), the interquartile range (25th and 75th percentiles, box), and the whiskers indicating the minimum and maximum values. Statistical analysis was performed using two-way repeated measures ANOVA followed by Sidak’s multiple comparison test (**p < 0.01, ***p < 0.001 and ****p < 0.0001; ns = not significant). P-EVOO promotes increased gene expression of key mRNAs potentially involved in NLRP3 inflammasome alternative pathway We next investigated whether dietary supplementation with P-EVOO modulates the expression of genes involved in NLRP3 inflammasome. To this end, mRNA expression levels of NLRP3, IL-1β, IL-18, GSDMD, CASP1, and CASP8 were evaluated in PBMCs isolated from obese children at baseline (T0) and after three months of nutritional intervention (T-END). Two-way repeated measures ANOVA revealed a significant effect of time on NLRP3 expression (p = 0.0001), with no significant effect of treatment group (p = 0.1153) and no significant time × group interaction (p = 0.0727), indicating an overall increase over time. However, post-hoc analysis using Šídák’s multiple comparisons test showed that this increase was primarily driven by the P-EVOO group, in which NLRP3 expression was significantly upregulated from T0 to T-END (p < 0.0001), while no significant change was observed in the R-OO group. In contrast, IL-1β expression showed a significant effect of time (p = 0.0002), group (p = 0.0231), and a significant time × group interaction (p = 0.0012), indicating a treatment-dependent modulation over time. Post-hoc analysis using Šídák’s multiple comparisons test demonstrated a marked increase in IL-1β expression in the P-EVOO group from T0 to T-END (p < 0.0001), which was not observed in the R-OO group, along with a significant difference between groups at T-END (p = 0.0002) (Fig. 7, Supplementary Table A5).Fig. 7Effects of polyphenol-enriched EVOO on the expression of inflammasome-related inflammatory markers. Relative mRNA expression levels of NLRP3, IL-18, IL-1β, GSDMD, CASP1, CASP8, PANX1 and P2RX7 in PBMCs from obese paediatric subjects of R-OO group (n = 13) and P-EVOO group (n = 17), measured before (T0) and after 3 months of intervention (T-END) with either refined olive oil (R-OO) or polyphenol-enriched EVOO (P-EVOO). mRNA levels were quantified by RT-qPCR from total RNA and normalized to GAPDH expression. Data are presented as relative ratio in mRNA expression and displayed as Tukey box-and-whisker plots. Statistical analysis was performed using two-way repeated measures ANOVA followed by Šídák’s multiple comparisons test. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns = not significant). Effects of polyphenol-enriched EVOO on the expression of inflammasome-related inflammatory markers. Relative mRNA expression levels of NLRP3, IL-18, IL-1β, GSDMD, CASP1, CASP8, PANX1 and P2RX7 in PBMCs from obese paediatric subjects of R-OO group (n = 13) and P-EVOO group (n = 17), measured before (T0) and after 3 months of intervention (T-END) with either refined olive oil (R-OO) or polyphenol-enriched EVOO (P-EVOO). mRNA levels were quantified by RT-qPCR from total RNA and normalized to GAPDH expression. Data are presented as relative ratio in mRNA expression and displayed as Tukey box-and-whisker plots. Statistical analysis was performed using two-way repeated measures ANOVA followed by Šídák’s multiple comparisons test. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns = not significant). Similarly, CASP1 expression was significantly affected by time (p = 0.0068), group (p = 0.0213), and their interaction (p = 0.0017), with Šídák post-hoc analysis confirming a significant increase in the P-EVOO group over time (p < 0.0001) and higher levels compared to R-OO at T-END (p = 0.0003). CASP8 expression also showed a significant interaction effect (p = 0.0488), with a significant increase over time observed in post-hoc analysis only in the P-EVOO group (p = 0.041), while no significant differences were detected in the R-OO group. No significant effects of time, group, or interaction were observed for IL-18 (time p = 0,14; group p = 0,30; interaction p = 0,06) and GSDMD (time p = 0,23; group p = 0,14; interaction p = 0,35), whose expression remained stable across the intervention in both groups (Supplementary Table A5). To further investigate whether increased mitochondrial ATP production could promote extracellular ATP release and subsequent K⁺ efflux, thereby triggering canonical NLRP3 inflammasome-related genes, we assessed the expression of PANX1 and P2RX7. No significant effects of time, group, or interaction were observed for either gene. Fold change analysis (T-END/T0) further supported these findings, revealing a significantly greater induction in the P-EVOO group compared to R-OO for NLRP3 (p = 0.0136), IL-1β (p = 0.0052), CASP1 (p < 0.0001), and CASP8 (p = 0.0471) (Supplementary Fig. B and Table A6). Notably, IL-18 displayed a significantly lower fold change in the P-EVOO group compared to R-OO (p = 0.0094), indicating a relative attenuation of IL-18 expression following P-EVOO supplementation. No significant differences between groups were observed for GSDMD (p = 0.7182), PANX1 (p = 0.6899), and P2RX7 (p = 0.861) (Supplementary Fig. B and Table A6). Overall, the observed transcriptional profile, characterized by increased expression of NLRP3, IL-1β, CASP1, and CASP8, in the absence of concomitant changes in GSDMD and IL-18, indicates a predominantly transcriptional activation of alternative inflammasome components, rather than activation of the canonical pyroptotic pathway. Discussion Childhood obesity is increasingly recognised as a condition characterised by marked metabolic and immunological disruption. Immune-cell function is closely shaped by metabolic reprogramming, with obesity driving a shift toward glycolysis and pro-inflammatory phenotypes21. Glycolysis and fatty-acid synthesis typically support inflammatory immune subsets such as M1 macrophages and effector T cells, whereas fatty-acid oxidation and oxidative phosphorylation support anti-inflammatory M2 macrophages and regulatory T cells6. In children with obesity, bioenergetic remodelling has been documented across multiple immune subsets8,9,22. The centrality of mitochondria in immune differentiation and activation further underscores their importance as regulatory players within obesity-related inflammation23. The NLRP3 inflammasome is a key sensor of metabolic stress and is activated in adipose-tissue immune cells, as well as in PBMCs of children with obesity24. Activation by saturated fatty acids, cholesterol crystals, and adipocyte-derived DAMPs leads to caspase-1–dependent IL-1β and IL-18 maturation, contributing to systemic inflammation, insulin resistance, and metabolic dysfunction in paediatric obesity24. Dietary patterns that mitigate chronic inflammation, particularly the MD and EVOO, have been shown to be beneficial in children with obesity11. EVOO exerts functional effects derived from both monounsaturated fatty acids and phenolic compounds17, including hydroxytyrosol, oleuropein, and tyrosol25. EVOO is particularly rich in monounsaturated fatty acids (MUFAs), primarily oleic acid, which have been consistently associated with improved lipid profiles, including reduced LDL cholesterol and increased HDL cholesterol26. They have also been linked to improved glycaemic control and lower blood pressure27. These monounsaturated fats modulate membrane fluidity and cell signalling, reduce oxidative stress, and support endothelial function by enhancing nitric oxide bioavailability, thereby attenuating chronic low-grade inflammation and cardiometabolic risk27. EVOO polyphenols exert potent antioxidant activity by scavenging reactive oxygen species and reducing oxidative stress, while also inhibiting pro-inflammatory signalling pathways, such as NF-κB, and promoting cytoprotective mechanisms, including Nrf2 activation28. In addition, these compounds have been shown to modulate lipid and glucose metabolism and improve endothelial function29. Together, the synergistic actions of EVOO MUFAs and polyphenols contribute to the attenuation of systemic inflammation and metabolic dysregulation associated with paediatric obesity. In our cohort, adherence to the MD, combined with either R-OO or P-EVOO, was associated with improvements in anthropometric and metabolic parameters, in line with previous evidence on the metabolic benefits of the MD11. However, a reduction in HOMA-IR was observed only in participants receiving P-EVOO supplementation. This effect may be partly explained by the potential ability of polyphenols to modulate oxidative stress and chronic low-grade inflammation, both of which are implicated in the development of insulin resistance. In addition, experimental evidence suggests that polyphenols may influence glucose metabolism by enhancing insulin-mediated glucose uptake in peripheral tissues and improving hepatic insulin sensitivity. Although findings from clinical studies and animal models indicate that the polyphenol content of extra-virgin olive oil may contribute to improvements in HOMA-IR, these effects cannot be fully separated from those of the Mediterranean diet as a whole or from the monounsaturated fatty acid content of olive oil29,30. Consistent with previous reports demonstrating links between mitochondrial respiration and insulin sensitivity31, P-EVOO intake improved mitochondrial functionality. PBMCs from the P-EVOO group showed higher coupled respiration, as well as maximal respiratory capacity, and increased spare respiratory capacity, indicating enhanced metabolic flexibility. These findings align with studies showing EVOO-related improvements in mitochondrial electron transport and ATP generation32,33. P-EVOO also increased glycolytic capacity and glycolytic reserve, supporting a coordinated enhancement of both mitochondrial and glycolytic pathways. Despite these adaptations, PBMCs from P-EVOO subjects exhibited a higher OCR/ECAR ratio, suggesting a predominant reliance on OXPHOS, typical of quiescent or regulatory immune states6. By contrast, PBMCs from the R-OO group showed no major metabolic changes. Even if in the PBMCs of the R-OO group we find an increase in ATP production, this could be likely to reflect the increased availability of fatty acids delivered by the oil, which may have served as additional substrates for β-oxidation17,34. Elevated mitochondrial ATP levels in the P-EVOO group further suggest improved respiratory efficiency associated with possible polyphenol-induced mitochondrial biogenesis35–38. Supporting this interpretation, P-EVOO intake is associated with upregulation of genes key regulators of mitochondrial biogenesis. Conversely, R-OO intake downregulated PPARGC1A and NRF1. Polyphenols are known to activate mitochondrial signalling pathways, including those related to PPARγ, AMPK and PKA39–41. Since oxidative stress is increased in paediatric obesity42, the improvement in antioxidant response represents a remarkable aspect. Antioxidant enzymes tend to decline with chronic obesity43,44, although some studies report compensatory upregulation45. In our study, P-EVOO promoted NFE2L2 and SOD2 expression, whereas R-OO decreased GPX1. These findings reinforce the hypothesis that polyphenols enhance mitochondrial function and antioxidant defences. A novel aspect of this work concerns potential inflammasome transcriptional profile regulation. In PBMC of the P-EVOO group, we observed increased gene expression of NLRP3, IL-1β, CASP8, and CASP1 without induction of IL-18 or GSDMD. This profile, although evaluated only at transcriptional level, might be incompatible with canonical or non-canonical inflammasome activation, both requiring GSDMD-mediated pyroptosis46,47. This transcriptional profile could instead be consistent with, the alternative NLRP3 inflammasome pathway, initially described in human monocytes48, although protein level assessment is needed to support this hypothesis. This pathway operates independently of potassium efflux and bypasses GSDMD pore formation, promoting sustained IL-1β secretion through TLR4–TRIF–RIPK1–FADD–caspase-8 signalling48,49. Caspase-8 emerges as a key regulator, enabling IL-1β maturation without pyroptosis. Speculative additional support for the transcriptional upregulation of inflammasome- alternative-pathway, derives from the absence of upregulation of pannexin-1 (Panx1) and P2RX7 mRNA levels, both essential for canonical and non-canonical NLRP3 signalling through ATP-mediated K+ efflux50–52. Notably, P2RX7 activation also promotes glycolytic reprogramming and Th1 differentiation53 whereas our P-EVOO PBMCs displayed a shift toward OXPHOS, further supporting a distinct activation modality. The independence of TLR4 signalling from the Panx1–ATP–P2RX7 axis54 is consistent with engagement of the alternative pathway, even if the functional significance of this mechanism requires further investigation and protein level assessment. One plausible driver of this response could be the release of circulating free fatty acids (FFAs) induced by polyphenol-mediated metabolic remodelling. EVOO stimulates lipolysis and increases FFA release55. Saturated FFAs, especially palmitic acid, are endogenous TLR4 ligands56 and may trigger low-grade TLR4-dependent activation of NLRP3 via the alternative pathway, enabling controlled IL-1β secretion without pyroptosis or IL-18 release. This “tolerated activation” could represent a physiological immune-metabolic adaptation. However, an important limitation of this study is that the analyses were restricted to gene expression data only, without the inclusion of protein-level measurements or functional assays (such IL-1β secretion), which would have provided more direct evidence of the activation of the underlying biological processes, thus no conclusions can be drawn regarding inflammasome activation or pathway engagement. Taken together, our findings suggest an integrated immuno-metabolic framework in which P-EVOO supplementation is associated with changes in the metabolism–mitochondria–immunity axis in children with obesity. The observed improvement in insulin sensitivity, occurring alongside enhanced mitochondrial respiration and increased antioxidant defences in PBMCs, is consistent with a potential role of EVOO polyphenols in modulating immune cell bioenergetics. Rather than reflecting a generic anti-inflammatory effect, these changes may indicate an association with shifts in cellular metabolic activity. The tendency toward a more oxidative metabolic profile, typically linked to quiescent or regulatory immune states, aligns with improved metabolic flexibility and increased antioxidants response—both relevant features of insulin resistance in paediatric obesity. Within this context, the speculative activation of components of the alternative NLRP3 inflammasome pathway may represent a possible immune-metabolic sensing response associated with metabolic adaptation. This non-pyroptotic, caspase-8– and caspase-1–dependent signalling could reflect a regulated inflammatory state that supports immune homeostasis rather than sustaining chronic low-grade inflammation. Limitations We acknowledge that isolating the specific effects of EVOO polyphenols within the context of a MD represents an inherent challenge, reflecting the real-world complexity of nutritional interventions in which multiple nutrients and bioactive compounds act synergistically. Although the use of refined olive oil as a comparator—sharing a similar fatty acid composition but lacking the phenolic fraction—provides a useful framework to explore the potential contribution of EVOO polyphenols, the specific attribution of the observed between-group differences, particularly in insulin resistance and cellular bioenergetics, cannot be definitively established within the present study design. Accordingly, the results should be interpreted cautiously within the context of dietary complexity and methodological constraints. Exploratory mechanistic analyses in PBMCs, including bioenergetic and gene expression profiling, were conducted in a relatively small subcohort selected based on sample availability and quality, and should therefore be considered exploratory. Larger, dedicated mechanistic studies are needed to confirm and extend these findings which will also need to be supported by the assessment of the protein level of the key components of the inflammasome pathway. The inclusion of children and adolescents without stratification by pubertal stage represents another limitation, as puberty may influence metabolic outcomes. Due to the limited sample size, stratified analyses were not feasible, and potential differences across pubertal stages cannot be excluded. Despite these limitations, the consistency of the observed effects supports the robustness of the main findings. Future studies with larger cohorts, pubertal stratification, objective measures of adherence, and dedicated mechanistic designs are warranted to further clarify the contribution of individual dietary components, including EVOO polyphenols. Conclusions Overall, while our data indicate associations between P-EVOO supplementation, mitochondrial function, antioxidant capacity, and modulation of NLRP3 inflammasome transcriptomic profile, causality cannot be inferred. Nevertheless, these observations support the hypothesis that polyphenol-rich nutritional interventions may contribute to improved immuno-metabolic balance in paediatric obesity. Larger, mechanistic, and longitudinal studies are required to confirm these findings and to clarify the underlying biological pathways. Methods Population enrolled This prospective, double-blind, randomized controlled study included 84 obese children and adolescents (BMI > 95th percentile) consecutively enrolled between January 2023 and December 2024 at the Endocrinology Unit of the Pediatric Hospital Giovanni XXIII, Aldo Moro University of Bari. Exclusion criteria were type 2 diabetes, secondary or syndromic obesity, hypothyroidism, Cushing’s disease, inflammatory diseases, metabolic or genetic disorders, and ongoing therapies for chronic systemic diseases. The study was conducted in accordance with the Declaration of Helsinki and was reviewed and approved by the Independent Ethics Committee of the Azienda Ospedaliera-Universitaria “Consorziale Policlinico” of Bari (protocol number 0062985/19/07/2021). Written informed consent to participate in the study was obtained from the participants’ parents or legal guardians. Intervention protocol Participants were randomly assigned in a 1:1 ratio to one of the two intervention groups: MD plus P-EVOO or MD plus R-OO. At enrollment, all subjects underwent a nutritional interview and received standardized instructions on the principles of the MD. Participants were instructed to consume the assigned oil for 12 weeks at a daily dose of 20 g, preferably in its raw form, replacing the oil normally used for meal preparation. Adherence to the intervention was monitored through structured telephone interviews conducted every 4 weeks. All participants were evaluated at baseline (T0) and after 12 weeks of oil supplementation (T-END). Dietary habits The adherence to the MD was evaluated using the KIDMED questionnaire, a validated tools for children and adolescents. The adherence is defined “poor” if the score is ≤ 3, average if score is between 4 and 7 and good if score is ≥ 857. Olive oil samples Two different types of olive oil (P-EVOO and R-OO) were provided by AGRIDÉ S.r.l. (Strada Vicinale Bartolomeo Chiarulli, 20, 70032 Bitonto, Bari, Italy) for use in the dietary intervention study. The two oils differed exclusively in their polyphenolic content (Table B, Supplementary material) while maintaining a similar fatty acid composition (Table C, Supplementary material). This was a crucial criterion to ensure that any changes observed in the metabolic or inflammatory parameters of the participants could be attributed solely to the different concentrations of polyphenolic compounds. Specifically, the following samples were used: P-EVOO obtained from blend of cultivars, characterized by the phenolic composition reported in Table A13; R-OO produced through a standard industrial refining process that removes the polyphenolic fraction, resulting in an oil composed almost exclusively of triglycerides and free fatty acids, while maintaining the same fatty acid profile as the EVOO. Both oil batches were bottled at the AGRIDÉ industrial plant in compliance with current EU regulations on food-grade packaging. The oils were filled into dark green glass bottles (1000 mL), labeled respectively with the letters A (R-OO) and B (P-EVOO) and randomly assigned in double-blind to each patient. To ensure product stability and prevent alterations during storage and transport, the bottles were packed in corrugated cardboard boxes designed to provide mechanical protection, thermal insulation, and light shielding. All samples were stored at ambient temperature (18–22 °C) in a dark, dry environment until distribution to the study participants. Clinical and anthropometric assessment A general clinical examination was performed in all participants. Anthropometric measurements included height (cm) and weight (kg). Body mass index (BMI) was calculated as weight divided by height squared (kg/m²) and expressed as a standard deviation score (SDS) based on the Italian BMI reference standards58. Pubertal development was assessed according to Tanner staging59. Waist circumference was also measured. Laboratory data assessment Blood glucose, insulin, total cholesterol (TC), high (HDL), low-density lipoprotein (LDL) cholesterol, triglycerides (TG), aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyl transferase (GGT), uric acid, IL1-β, and IL6, C reactive protein, were measured after overnight fasting in all subjects at T0 and at T-END. TC, LDL, HDL, and TG values were considered in the normal range within the 5th and 95th percentiles. Insulin resistance was measured by the homeostasis model assessment (HOMA) index, calculated as insulin (µU/mL) × blood glucose (mmol/L)/22.5. Bioenergetic analysis Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured as key bioenergetic parameters of PBMCs using an XF24 Extracellular Flux Analyzer (Seahorse Bioscience, Billerica, MA, USA) following the manufacturer’s instructions. Briefly, after optimization of the seeding density, 3.5 × 105 PBMCs were seeded in XFe24 plates coated with poly-D-lysine and incubated until complete adhesion in XF RPMI medium supplemented with glucose 10mM, L-glutamine 2mM, and sodium pyruvate 1mM in a non-CO2 incubator at 37 °C. OCR and ECAR were measured in intervals by mixing and waiting periods, using a set number of replicates per experimental point: basal OCR measurements were followed by oligomycin injection (1µM) to inhibit ATP synthase followed by sequential injections of 2µM Carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP) to uncouple mitochondrial respiration, rotenone + antimycin A (1µM + 1µM) to inhibit mitochondrial respiration. For ECAR analysis, at the end of running, 2-deoxy-d-glucose 50mM was injected to shut down glycolysis and allow data correction for non-glycolytic medium acidification. The OCR and ECAR values were normalized to protein content in each well, determined by the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Final mitochondrial parameters were calculated as an average of four replicate as follows: basal respiration corresponds to OCR measurement before oligomycin injection; coupled respiration is the difference between basal respiration and OCR measurement after oligomycin injection; maximal respiration is the last OCR measurement after FCCP injection and spare capacity corresponds to the difference between last OCR measurement after FCCP injection and basal respiration. All mitochondrial parameters were corrected by subtracting non-mitochondrial respiration (last measurement after rotenone/antimycin injection). Proton leak corresponds OCR measurements after oligomycin injection, subtracting non-mitochondrial respiration. Final ECAR parameters were calculated as follows: glycolysis is the basal measurement before 2-deoxyglucose injection; glycolytic capacity corresponds to the difference between measurements after oligomycin and 2-deoxyglucose injections; glycolytic reserve is calculated as the difference between measurements after oligomycin injection and basal measurement. All glycolytic parameters were corrected by subtracting last measurement after 2-deoxyglucose injection (non-glycolytic acidification). The cellular ATP production rates were quantified using the Agilent Seahorse XF ATP real-time assay kit using label-free technology in real time as specified by the manufacturer and essentially as described in19. According to the equation: glucose + 2 ADP + 2 Pi → 2 lactate + 2 ATP + 2 H2O + 2 H+, the glycolytic ATP production rate (GlycoATP) was estimated from glycolytic proton production rate (PPRglyc) after determination of the buffering power (BP). BP was determined separately by measuring the pH changes elicited by successive additions of titrated HCl in XF assay media in the presence of the same number of assayed cells and was found to be 3.1 mmol H+/L/pH. The mitochondrial ATP production rate (MitoATP) was equal to OCRATP×2×P/O (mol ATP/mol O), where OCRATP was the difference between basal OCR and OCR in the presence of oligomycin. A theoretical P/O ratio of 2.25 was assumed, based on the combined oxidation of glucose, fatty acid, and glutamine. RNA isolation, reverse transcription and quantitative PCR Total RNA from cells was isolated using the miRNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. Complementary DNA (cDNA) was synthesized by reverse-transcription of total RNA using the iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA, USA), following the manufacturer’s instructions. Semi-quantitative determination of mRNA levels was performed by Real-Time Quantitative Reverse Transcription PCR (qRT-PCR) using the SsoAdvanced Universal SYBR® Green Supermix (Bio-Rad, Hercules, CA, USA). Reactions were performed in a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA) in triplicate for each sample for three independent experiments. Relative quantification was performed using the comparative CT method (ΔΔCT). Quantitative normalization for each sample was performed by using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as an internal control. Validated primers for semi-qRT-PCR are provided in Table D (Supplementary material). Statistical analysis Data are reported as mean ± standard error mean (SEM). The Shapiro–Wilk normality test was applied for each data set to determine if the data had a normal distribution. Statistical analyses were performed using two-way repeated measures ANOVA or mixed-effects models (REML), with Sidak’s correction for multiple comparisons. The threshold for statistical significance was set to 0.05. All analyses were performed using GraphPad Prism Software Version 9 (GraphPad Software, San Diego, CA, USA). Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1