Effects of the Mediterranean diet on oxidative stress: a systematic review and meta-analysis of randomized clinical trials

Abstract

Background

The Mediterranean diet (MD) is rich in antioxidant components that can increase the antioxidant capacity of the body and improve the oxidative stress status. This systematic review and meta-analysis aimed to determine the ability of the MD to regulate oxidative stress status.

Methods

This study was conducted in accordance with the PRISMA guidelines. The PubMed, Embase, Cochrane Library, Web of Science, ClinicalTrials, and Chinese Clinical Trial databases were searched from their inception to November 2024. Randomized controlled trials that investigated the relationship between MD patterns and oxidative stress biomarkers were selected to assess the impact of the MD on oxidative stress markers.

Results

The 20 studies included in the analysis encompassed 2,208 participants and examined 13 indicators. The MD increased the total blood antioxidant capacity (standardized mean difference [SMD]: 0.57, 95% confidence interval [CI]: 0.31 to 0.83; p < 0.001), decreased urine F2-isoprostanes (SMD: -0.40, 95% CI: -0.67 to -0.12; p = 0.005), and significantly decreased blood malondialdehyde levels (SMD: -0.70, 95% CI: -1.12 to -0.27; p = 0.001).

Conclusion

These findings suggest that the MD can enhance the body’s antioxidant capacity and improve lipid peroxidation. Therefore, an MD may be an effective method for reducing oxidative stress.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-026-08233-8.

Introduction

Oxidative stress refers to the disproportionate production of oxidants and antioxidant defenses in favor of oxidants, resulting in disrupted reduction-oxidation (redox) signaling and potential molecular damage [1, 2]. High pro-oxidant activity induces oxidative damage to cellular components, such as lipids, DNA, and proteins, thereby destroying cellular structures, interfering with signaling pathways, promoting inflammatory responses, and inducing gene mutations. Various byproducts generated during this process are usually used as indicators to evaluate the oxidative stress status. These biomarkers can be classified into two categories [3, 4]. One category comprises products that arise from the reaction of DNA, lipids, proteins, and other molecules with reactive oxygen species (ROS) or reactive nitrogen species in the cellular microenvironment. These products are commonly used as indicators of oxidative damage [1]. Such biomarkers include products of lipid peroxidation such as thiobarbituric acid reactive substances (TBARS), oxidized low-density lipoprotein (OX-LDL), malondialdehyde (MDA), and F2-isoprostanes (F2-IsoP). Biomarkers of oxidative DNA damage include 8-hydroxy-2′-deoxyguanosine (8-OHdG) [5], whereas nitrotyrosine is an indicator of protein oxidative damage [6]. Another category includes changes in the antioxidant system, such as total antioxidant capacity (TAC) and ferric-reducing antioxidant power (FRAP), which reflect the antioxidant capacity of the body. Additionally, changes in the levels of antioxidants, including paraoxonase-1 (PON-1), catalase (CAT), glutathione (GSH), superoxide dismutase (SOD), and glutathione peroxidase (GPX), indicate antioxidant function [3]. Oxidative stress is a key pathological mechanism in some chronic diseases, including cardiovascular diseases [7], neurological disorders [8], inflammatory bowel disease [9], cancer [7, 10], metabolic syndrome [6], and lung diseases [11–14]. Oxidative stress plays a key role in different phases of human development and aging. Therefore, maintaining health by regulating oxidative stress in the body is important.

The Mediterranean diet (MD) is a healthy eating pattern rich in foods containing highly unsaturated fats. High intakes of legumes, cereals, fruits, vegetables, nuts, and fish, along with the predominant use of olive oil as a dietary fat source, are the defining features of the MD. This pattern also incorporates a moderate intake of dairy products (such as cheese and yogurt), low consumption of other meats, and appropriate consumption of red wine [15, 16]. The MD provides additional antioxidants, and its high polyphenol content helps reduce ROS production. It is also abundant in monounsaturated fatty acids, carotenoids, selenium, and vitamins C and E, among other antioxidants, which can improve the body’s oxidative stress status. Moreover, the synergistic effects of some antioxidants in the MD may surpass their individual effects [17]. For example, the combination of carotenoids and vitamin E inhibits lipid peroxidation more effectively than either of these alone [18]. As a non-drug intervention, the MD has significant effects on chronic diseases, including obesity, type 2 diabetes [19], and nonalcoholic fatty liver disease (NAFLD) [20]. Furthermore, the MD has protective effects against cognitive decline and thyroid autoimmune disorders [21, 22].

In patients with NAFLD aged 11–18 years, the MD did not lead to significant alterations in oxidative stress-related biomarkers, except for GSH, when compared with a low-fat diet [23]. Sofi et al. reported that the MD improved circulating cardiovascular risk parameters in patients by attenuating oxidative stress [24]. Al-Aubaidy et al. demonstrated that an MD intervention reduced the levels of the DNA oxidative damage marker 8-OHdG [25], contrary to observations reported by Georgoulis et al. [26]. Although some studies have reported that an MD has antioxidant effects, previous studies have often analyzed the MD in combination with other nutritional supplements, and only quantified a limited number of oxidative stress markers [27, 28]. Given these factors, we conducted a systematic review and meta-analysis to comprehensively evaluate the oxidative stress markers in patients who received an MD as the sole intervention.

Methods

This present systematic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [29] and was registered in the International Prospective Register of Systematic Reviews (PROSPERO) (CRD420250653823).

Information sources and search strategy

We searched the Chinese Clinical Trial Registry, Cochrane Library, Embase, ClinicalTrials.gov, Web of Science, and PubMed, from database inception to November, 2024, as described by Razieh Pirouzeh et al. [30]. We used both Medical Subject Headings (MeSH) terms and free-text words. The intervention search strategy included “Diet, Mediterranean” (MeSH term), replacing the oxidative stress marker “8-oxo-7-hydro-deoxyguanosine” with “8-hydroxy-2’-deoxyguanosine” (MeSH term), and “randomized controlled trials” (MeSH term) as the study design. Additionally, we refined the search strategy for F2-IsoP to ensure comprehensive coverage of the literature. Finally, the MeSH terms for intervention measures, oxidative stress indicators, and study design were connected using the Boolean term “AND” while the free text words were connected using “OR.” In the databases of the Chinese Clinical Trial Registry, the use of “Mediterranean Diet” or “Mediterranean” as the primary subject requires the publication of the results documents. The intervention was based on “Diet, Mediterranean.” We also searched ClinicalTrials.gov for studies with available results. According to the prespecified eligibility criteria, two authors independently evaluated the retrieved studies, with any disagreements settled by a third party. The comprehensive search strategies are listed in Additional file 1.

Inclusion and exclusion criteria

The included studies were parallel or crossover randomized controlled trials that investigated the impact of the MD on oxidative stress in humans, with no restrictions on the study population. Only studies that included at least one of the following markers were considered: TAC, OX-LDL, SOD, MDA, GSH, GPX, 8-OHdG, FRAP, PON-1, CAT, TBARS, F2-IsoP, and nitrotyrosine. Studies were allocated to different meta-analyses according to the oxidative stress biomarkers reported. If the marker levels were assessed at multiple time intervals within a study, the longest follow-up results were selected to capture the prolonged effects of the intervention. The intervention in the experimental group consisted only of the MD, without additional nutritional or food supplements, or any extra components of the MD. However, if the control group received the same additional interventions, it was still included. The dietary interventions in the control group were not restricted. Clinical trial registration platforms require the publication of experimental data. The shortest intervention duration was 2 weeks, and studies focusing on postprandial changes in oxidative stress markers were excluded. Cohort studies, nonrandomized controlled trials, scross-sectional studies, case-control studies, and studies using other supplements or additional interventions were also excluded.

Quality assessment

The risk of bias was assessed using the Cochrane Risk of Bias Tool (version 1.0) [31], which evaluates seven aspects: selective reporting, random sequence generation, incomplete outcome data, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, and other potential sources of bias. For each aspect, the risk of bias was classified as low, high, or unclear. If at least three aspects were rated as low risk, the study was considered high quality, while studies with at least two and one aspect rated as low risk were considered moderate- and low-quality studies, respectively [20]. As the intervention involved dietary changes, blinding of the participants was difficult. Therefore, this assessment was based on other factors. Two reviewers independently assessed the retrieved studies, and any disagreements were resolved by a third party.

Data extraction

We created data collection sheets using Microsoft Excel, and data were independently extracted by two researchers, with a third party consulted to settle any disagreements. We extracted the authors names, publication date, country, study design, outcome measurement methods, intervention duration, sample size, participant disease, age, sex, body mass index (BMI), and the mean and standard deviation (SD) or median and interquartile range (Q1 and Q3) of each oxidative stress indicator. Before pooling the effect sizes, all data were standardized to the mean ± SD.

Statistical analysis

Review Manager (version 5.3.) was used to analyze the extracted data. When data for the same oxidative stress indicator were available from at least two studies, a meta-analysis was performed to determine the effect size. For continuous outcome variables assessed using the same measurement method, the mean difference (MD) was calculated using the inverse variance approach. For studies that used different measurement methods or units, the standardized mean difference (SMD) was used as the effect size [32, 33]. We categorized the effect sizes for SMD as described by Zhang.et al. [34] (small effect: SMD < 0.40; moderate effect: SMD 0.40–0.70; large effect: SMD > 0.70). Before combining the effect sizes, we assessed the heterogeneity among the studies using the Q test and I² statistic. Heterogeneity was considered negligible when p > 0.1 and I² < 50%, in which case a fixed-effects model was applied. Otherwise, a random-effects model was used [35]. As fewer than 10 studies were available for each indicator, we did not assess publication bias [36]. Sensitivity analyses were conducted by sequentially removing one study at a time to assess the influence of individual studies on the overall effect size. Statistical significance was set at p < 0.05.

Results

Study selection

We identified 233 studies in PubMed, 328 in the Cochrane Library, 1,031 in Embase, and 997 in the Web of Science. Additionally, 13 clinical trials with results were found on ClinicalTrials.gov. No studies were found in the Chinese Clinical Trial Registry database. A total of 2,602 studies were screened, and 743 duplicates were excluded. Among the remaining studies, 639 were excluded because they were animal studies, reviews, systematic reviews, or meta-analyses. After reading the titles and abstracts of the remaining studies 995 studies inconsistent with the research focus were excluded. The full texts of 10 articles were inaccessible. The remaining 215 articles were screened through a full-text review. Studies were excluded from the quantitative analysis because of unclear reporting of specific change values or the inability to extract useful data on the oxidative stress markers. After removing studies with unrelated interventions, research designs, outcome variables, or content, as well as reviews, letters, funding projects, and study protocols, the final analysis included 20 studies. The study selection process is illustrated in Fig. 1.

Fig. 1.

The specific study selection process

Characteristics of the included studies

The studies included in the analysis were published between 2004 and 2024 and involved 2,208 participants from 13 countries, mainly in Europe. The total sample sizes ranged from 20 to 1,002 participants. The experimental and control groups included 994 and 933 participants, respectively. Among the 20 included randomized controlled trials, three used a crossover design method, three investigated the effect of the MD on TAC, two on PON-1, three on TBARS, two on MDA, two on GSH, two on GPX, two on SOD, six on F2-IsoP, two on FRAP, two on 8-OHdG, two on OX-LDL, two on CAT, and two on nitrotyrosine. The shortest MD intervention lasted 14 days, and most control groups followed a low-fat, low-calorie, or saturated fatty acids rich diet. Only one study included adolescents, whereas the others included adults. Two studies included only women, one included only men, and the remaining studies included both sexes. The selected studies involved healthy populations and groups with obesity, NAFLD, kidney disease, metabolic syndrome, cardiovascular disease, cancer, Parkinson’s disease, obstructive sleep apnea, and type 2 diabetes. The meta-analysis included 20 studies [23, 25, 26, 37–53]. An overview of the study characteristics are provided in Table 1.

Table 1.

The characteristics of the studies included in this research

Author	Time	Country	Study design	Paticipants’ characteristics	Study Duration	N		Diet type		Sex (male/female)		Age(years, Mean ± SD)		BMI(kg/m2,Mean ± SD)
						Intervention	Control	Intervention	Control	Intervention	Control	Intervention	Control	Intervention	Control
Al-Aubaidy.et al	2021	Australia	crossover	Participants with well-controlled T2DM.	24 w	19	19	Mediterranean	Habitual diet	11/8	11/8	57.50 ± 1.30	57.50 ± 1.30	31.54 ± 1.19	31.54 ± 1.19
Ambring.et al	2004	Sweden	crossover	Characteristics covering 95% of Sweden’s adult population.	4 w	22	22	Mediterranean	Swedish diet	12/10	12/10	43.00 ± 1.00	43.00 ± 1.00	26.00 ± 0.60	NA
Barona.et al	2012	United States	parallel	Women with metabolic syndrome and high plasma concentrations of LDL cholesterol (LDL-C ≥ 100 mg/dL).	12 w	15	20	Mediterranean	Medical food	NA	NA	49.40 ± 9.20	44.00 ± 10.30	33.70 ± 5.00	32.90 ± 5.60
Buscemi.et al.	2009	Italy	parallel	The age ranged from 30 to 50 years and a range of BMI: 27.00–39.90 kg/m2. overweight-obese otherwise healthy, non-smoking and non-pregnant women participants.	2 m	10	10	Mediterranean	Atkins low- carbohydrate diet	0/10	0/10	39.00 ± 3.00	38.00 ± 3.00	34.00 ± 1.00	34.50 ± 1.80
Ceriello.et al.	2014	Spain	parallel	T2DM patients.	3 m	12	12	Mediterranean	Low-fat diet	9/3	8/4	NA	NA	29.80 ± 1.40	29.20 ± 1.10
Davis.et al	2017	Australia	parallel	Older adults aged > 65 years, nonsmokers, and free of chronic disease.	6 m	70	67	Mediterranean	Habitual diet	36/49	37/44	71.00 ± 4.90	70.80 ± 4.70	26.70 ± 3.70	27.40 ± 4.10
Gutierrez-Mariscal.et al	2024	Spain	parallel	Patients aged 59.5 ± 0.3 years, who had established CHD and no severe illness, with a BMI of 31.1 ± 0.1.	5 y	432	377	Mediterranean	Low-fat diet	NA	NA	NA	NA	NA	NA
Georgoulis.et al	2021	Greece	parallel	Adults aged 18–65 years, overweight or obese, BMI ≥ 25 kg/m2, otherwise healthy individuals with moderate or severe OSA (AHI) ≥ 15 events/h of sleep.	6 m	59	62	Mediterranean	Low-calorie diet	44/18	51/14	50.0 ± 9.10	48.00 ± 10.00	NA	NA
Harvie.et.al	2021	United Kingdom	parallel	Breast cancer participants with age ≥ 18 years, hemoglobin > 110 g/L, and BMI > 19 kg/m².	5 m	85	84	Mediterranean	Intermittent energy restriction diet	0/85	0/84	52.60 (24.00–77.00)#	51.20 (31.00–71.00)#	28.20 ± 6.10	28.00 ± 6.20
Hadziabdic.et al	2015	Croatia	parallel	Eligible patients were 47.5 ± 12.4 years old, with a BMI of 41.59 ± 7.32.	1 y	40	44	Mediterranean	Standard hypolipemic diet	19/44*	13/48*	46.20 ± 12.70*	49.00 ± 12.10*	41.98 ± 7.59*	41.20 ± 7.07*
Jaacks.et al	2018	United States	parallel	Participants aged 40–65 years, BMI ≥ 28 kg/m2 and < 35 kg/m2,nonsmokers, currently con- suming a stable habitual diet not consuming anti-oxi- dants or vitamin-mineral preparations in the past 4 weeks.	8 w	11	9	Mediterranean	Habitual high-fat American-type diet	NA	NA	NA	NA	NA	NA
Kolomvotsou.et al	2013	Greece	parallel	Healthy employees with abdominal obesity (waist circumfer- ence > 102 cm for men and > 88 cm for women).	2 m	41	41	Mediterranean	Not specified	23/18	20/21	50.20 ± 6.40	50.60 ± 8.10	31.50 ± 3.80	32.80 ± 4.70
Mayr.et al	2019	Australia	parallel	Adults with CHD	6 m	34	31	Mediterranean	Low-fat diet	27/7	27/4	61.80 ± 9.20	61.80 ± 9.50	30.70 ± 5.00	29.10 ± 5.30
Marin.et al	2011	Spain	crossover	Free-living elderly participants (aged > 65 years); none of the participants showed evidence of chronic illness.	4 w	20	20	Mediterranean	SFA-rich diet	10/10	10/10	67.10 ± 4.52	67.10 ± 4.52	31.90 ± 5.50	31.90 ± 5.50
Mekki.et al.	2010	Algeria	parallel	Patients aged 61 ± 14 years were included on the basis that they had a moderate CRF.	6 m	20	20	Mediterranean	NKFK/DOQI Guidelines Recommended Diet	11/9	10/10	60.00 ± 10.00	59.00 ± 12.00	NA	NA
Paknahad.et al	2020	Iran	parallel	Participants aged > 40 years with idiopathic Parkinson’s disease.	10 w	36	34	Mediterranean	Traditional Iranian Diet	20/16	21/13	59.30 ± 8.30	58.60 ± 9.30	25.80 ± 3.30	25.30 ± 2.70
Parcina.et al	2015	Germany	parallel	Males aged 20–40 years in good health (absence of acute or chronic disease).	2 w	14	13	Mediterranean	German cooking style diet	14/0	13/0	31.88 ± 6.25	29.14 ± 5.83	24.01 ± 1.53	24.54 ± 2.27
Papandreou.et al	2012	Greece	parallel	The patients were middle-aged, with an average age of 48.1 ± 12.4 years; The majority were men (80.96%) with moderate to severe OSAHS and a BMI of 36.6 ± 3.7 kg/m 2.	6 m	11	10	Mediterranean	Prudent diet	NA	NA	NA	NA	36.20 ± 3.3.	37.10 ± 4.40
Stachowska.et al	2005	Poland	parallel	Kidney graft recipients with stable graft function, nonsmokers.	6 m	21	16	Mediterranean	Low-fat diet	15/6	10/6	41.00 ± 12.50	46.00 ± 9.50	25.00 ± 4.10	26.20 ± 4.40
Yurtdas.et al	2021	Turkey	parallel	Participants’ ages ranged from 11 to 18 years, with a BMI ≥ 95th percentile, and diagnosed with grade ≥ 1 NAFLD.	12 w	22	22	Mediterranean	Low-fat diet	13/9	13/9	13.00 ± 1.99	13.90 ± 2.34	NA	NA

*: Data collected before any loss to follow-up occurred at the start of the study

#: Mean(range)

w, week; m, month; y, years; BMI, Body mass index; SFA, Saturated fatty acid; NKFK/DOQI, National kidney foundation-kidney disease outcomes quality initiative; CRF, chronic renal failure; CHD, coronary heart disease; NAFLD, nonalcoholic fatty liver disease; T₂DM, type 2 diabetes mellitus; AHI, apnea hypopnea index; OSA, obstructive sleep apnea; OSAHS, obstructive sleep apnoea/hypopnoea syndrome

Risk of bias

A summary of the risk of bias across the studies is presented in Fig. 2. Owing to the nature of the intervention, which involved changing dietary habits, blinding of the participants was difficult. Three studies reported blinding of participants, whereas all other studies presented either an unclear or high risk of bias. Among the 20 included studies, three mentioned allocation concealment, seven mentioned blinding of outcome assessment, and eleven provided sufficient information for the assessment of incomplete outcome data.

Fig. 2.

Risk assessment of the included studies

Summary of oxidative stress indicators

Indicators of oxidative stress can be categorized into two main categories. One category comprises biomarkers of oxidative damage, which include indicators such as TBARS, MDA, F2-IsoP, OX-LDL, 8-OHdG, and nitrotyrosine. The second category includes indicators of the antioxidant defense system, such as TAC, FRAP, CAT, GSH, PON-1, GPX, and SOD.

Biomarkers of oxidative damage

F2-IsoP

Six studies investigated the effects of the MD on F2-IsoP [26, 37–41]. Among the included studies, three assessed alterations in urine F2-IsoP levels after MD intervention [26, 37, 38]. Compared with the control diet, the meta-analysis showed a significant reduction in urine F2-IsoP levels in the MD group, with a moderate negative effect (SMD: -0.40, 95% CI: -0.67 to -0.12; p = 0.005). The results showed no heterogeneity (I² = 0%) (Fig. 3).

Fig. 3.

Effect of Mediterranean diet on urine F2-isoprostanes (F2-IsoP)

The results of a meta-analysis of three studies on the effect of dietary intervention on blood F2-IsoP levels [39–41] showed decreased levels in the MD group compared to that in the control diet group. However, this difference was not significant (SMD: -2.00, 95% CI: -4.30 to 0.30; p = 0.09). The three included studies showed significant heterogeneity (I² = 94%) (Supplementary Fig. 1A in Additional file 2). The sensitivity analysis revealed that the results of study [40] significantly affected the synthesized results; therefore, this study (the only study that included type 2 diabetes in addition to obesity) was excluded from the analysis, which improved the heterogeneity (I² = 31%, p = 0.44). However, the results remained non-significant (SMD: -0.12, 95% CI: -0.44 to -0.19; p = 0.44) (Supplementary Fig. 1B in Additional file 2).

MDA

Two studies described the effect of the MD on blood MDA [42, 43]. Compared with the control diet, the meta-analysis results showed a significant reduction in blood MDA levels in the MD group, with a moderate negative effect (SMD: -0.70, 95% CI: -1.12 to -0.27; p = 0.001). The results showed no heterogeneity (I² = 0%) (Fig. 4).

Fig. 4.

Effect of Mediterranean diet on blood Malondialdehyde (MDA)

TBARS

The results of the meta-analysis of three studies on the effect of the MD on blood TBARS levels [44–46] showed that the MD reduced urinary TBARS levels compared with the control diet, although this result was not significant (SMD: -2.57, 95% CI: -5.65 to 0.5; p = 0.10). The results showed significant heterogeneity (I² = 96%) (Supplementary Fig. 2A in Additional file 2).

OX-LDL

Two studies described the effect of MD on blood OX-LDL [37, 47]. However, the results of this meta-analysis indicated no significant effects (SMD: -2.35, 95% CI: -5.90 to 1.20; p = 0.19). The results showed significant heterogeneity (I² = 96%) (Supplementary Fig. 2B in Additional file 2).

Nitrotyrosine

Two studies involving 44 participants reported changes in blood nitrotyrosine levels [37, 40]. The findings showed that MD had no significant effect on nitrotyrosine levels compared to the control group (MD: -146.19, 95% CI: -426.95 to 134.57; p = 0.31). The results showed significant heterogeneity (I² = 100%) (Supplementary Fig. 2C in Additional file 2).

8-OHdG

The results of the meta-analysis of two studies that described the effects of the MD on blood 8-OHdG levels [25, 42] after 2 weeks showed that the MD reduced 8-OHDG levels compared to the control diet. However, this difference was not significant (MD: -12.37, 95% CI: -41.68 to 16.94; p = 0.41). The results showed significant heterogeneity (I² = 96%) (Supplementary Fig. 2D in Additional file 2).

Changes in the antioxidant system

TAC

Among the included studies, three [48–50]described the effects of the MD on blood TAC. The study population included obese individuals aged > 40 years with idiopathic Parkinson’s disease. The follow-up period for all studies was > 10 weeks. The meta-analysis results showed that, compared with the control group, the MD intervention significantly increased TAC with a moderate positive effect (SMD: 0.57, 95% CI: 0.31 to 0.83; p < 0.0001). The results showed moderate heterogeneity (I² = 45%) (Fig. 5).

Fig. 5.

Effect of Mediterranean diet on blood total antioxidant capacity (TAC)

GPX

The meta-analysis of two studies reporting changes in GPX levels [23, 37] showed no significant difference after the MD intervention compared with the control diet (SMD: -0.11, 95% CI: -0.54 to 0.32; p = 0.63). The combined results indicated moderate heterogeneity (I² = 44%) (Supplementary Fig. 3A in Additional file 2).

SOD

Two studies reported changes in SOD levels. A random-effects model was used to incorporate studies [23, 37] into the quantitative synthesis. The meta-analysis showed no significant difference after the MD intervention compared with the control diet (SMD: -0.34, 95% CI: -1.31 to 0.62; p = 0.49). The results showed significant heterogeneity (I² = 79%) (Supplementary Fig. 3B in Additional file 2).

CAT

Two studies [23, 37] involving 102 participants reported changes in blood CAT levels. The fixed-effects model showed that, compared with the control group on other diets, the MD decreased CAT levels in blood samples, but the difference was not significant (SMD: 0.04, 95% CI: -0.39 to 0.47; p = 0.87). The results showed moderate heterogeneity (I² = 32%) (Supplementary Fig. 3C in Additional file 2).

PON-1

The meta-analysis of two studies [23, 51] on the effects of the MD on blood PON-1 levels indicated no significant differences between the MD and control diet (SMD: -0.20, 95% CI: -0.47 to 0.07; p = 0.15). The heterogeneity in blood PON-1 levels among the studies was not significant (I² = 6%) (Supplementary Fig. 3D in Additional file 2).

GSH

The meta-analysis of two studies [52, 53] on the effects of the MD on blood GSH levels revealed no significant difference between the MD and control diet (SMD: 7.56, 95% CI: -8.55 to 23.66; p = 0.36). The heterogeneity in blood GSH levels among the studies was significant (I² = 100%) (Supplementary Fig. 3E in Additional file 2).

FRAP

The meta-analysis of two studies [40, 49] on the effects of the MD on FRAP showed no significant difference between the MD and control diet (SMD: 8.94, 95% CI: -7.70 to 25.57; p = 0.29). However, significant heterogeneity was observed between the studies (I² = 97%) (Supplementary Fig. 3F in Additional file 2).

Discussion

This systematic review summarized 20 articles from multiple countries that assessed the levels of 13 oxidative stress markers: TAC, F2-IsoP, MDA, TBARS, GPX, SOD, CAT, nitrotyrosine, 8-OHdG, OX-LDL, PON-1, FRAP, and GSH. The meta-analyses demonstrated that the MD had a significant impact on TAC, F2-IsoP, and MDA levels compared to control diets.

Oxidative stress can cause lipid peroxidation, leading to cellular dysfunction and tissue damage, and is an important factor in neurodegeneration [54], atherosclerosis [55], retinal injury [56], and other diseases [57]. As a non-pharmacological intervention, the MD may help reduce lipid peroxidation. Changes in 13 biomarkers may be relevant to oxidative stress-related diseases. The MD may ameliorate oxidative stress through various mechanisms. The polyphenols prevalent in this diet can reduce ROS production, thereby attenuating the activation of nuclear factor kappa-B (NF-κB). This process prevents the depletion of GSH and overproduction of ROS induced by its upregulation, while also enhancing the nuclear factor erythroid 2-related factor 2 (Nrf2)-mediated antioxidant pathways [58–60]. Resveratrol, a polyphenol found in the MD, regulates lipid metabolism by activating the AMP-activated protein kinase (AMPK) pathway. Furthermore, by modulating the sirtuin 1 (SIRT1), Nrf2, and NF-κB pathways, it improves the body’s antioxidant capacity and reduces polyunsaturated fatty acid oxidation. This results in changes in marker levels including increased CAT, SOD, GSH, and GPX levels, and reduced MDA levels, thereby improving oxidative stress [61, 62]. Carotenoids, which are abundant in the MD, can alter cell membrane properties, enhancing their ability to resist ROS. Additionally, vitamins such as vitamin E, found in the MD, can inhibit lipid peroxidation and exert antioxidant effects in hydrophobic environments.

High adherence to the MD significantly reduces the risk of mortality. Moreover, the synergistic effects of its various nutrients effectively improve health [63]. For example, although carotenoids and vitamin E can individually improve lipid peroxidation, their synergistic effect is superior, and they can jointly enhance the antioxidant activity of vitamin C [18, 64].

Our results showed that the MD significantly increased blood TAC levels compared to control diets, with a moderate positive effect on TAC. TAC reflects the overall antioxidant buffering capacity of the body rather than that of individual antioxidants. It is used to evaluate the cumulative and potentially synergistic effects of antioxidants [17, 65]. The increase in TAC suggests that the MD enhances the antioxidant capacity of the body and reduces oxidative damage. An observational study also reported similar results, demonstrating that the MD significantly increased the TAC [66]. As mentioned above, the MD has the characteristic that the synergistic beneficial effects of various dietary components are more pronounced. Therefore, the increase in TAC observed in this study supports the beneficial effect of the MD on overall antioxidant status.

In this study, the markers of lipid peroxidation, particularly F2-IsoP in urine, showed a significant decrease. F2-IsoP is a lipid peroxidation product of oxidative stress and is widely used as a biomarker of oxidative stress within the body. This decrease suggests that lipid peroxidation was inhibited. The production mechanism primarily involves the non-enzymatic peroxidation of arachidonic acid under free radical attack [67]. However, the blood F2-IsoP levels did not significantly decrease. This difference suggests that urinary but not blood F2-IsoP is a more suitable marker of oxidative stress [68]. Our results showing that F2-IsoP levels decreased after the MD intervention are consistent with previous reports [69]. The reduction in urinary F2-IsoP levels may be due to the antioxidant-rich composition of the MD, which provides phenolic compounds, vitamin E, and carotenoids. Among these, phenolics may contribute to the protection of lipids from oxidative stress [70], vitamin E can inhibit free radical-mediated lipid peroxidation [71], and carotenoids can quench lipid radicals and reduce lipid peroxide formation in biological membranes [72]. Together, these antioxidant components may neutralize free radicals, thereby contributing to lower F2-IsoP production.

We also found that another important marker of lipid peroxidation, MDA [73, 74], was significantly decreased after an MD. This finding is consistent with a previous report that greater adherence to the MD was associated with a lower MDA level [75]. Fruits, vegetables, nuts, and olive oil in the MD are rich in polyphenols, which can inhibit ROS-mediated oxidative stress [70, 76–78]. Furthermore, monounsaturated fatty acids abundant in the MD can reduce the susceptibility of lipids to oxidative stress [79]. Vitamin E and carotenoids, also abundant in this diet, can effectively scavenge lipid peroxyl radicals [79]. All these ingredients in the MD ultimately reduce MDA formation. TBARS is also a commonly used marker of lipid peroxidation. It does not refer to a single compound, but rather to a group of substances that react with TBA, among which MDA is the most representative. The meta-analysis results showed that MD reduced TBARS levels, although the differences were not statistically significant. This finding complements the MDA results, suggesting that the MD can help alleviate lipid peroxidation.

However, in this meta-analysis, no significant changes were observed in the levels of GPX, SOD, CAT, nitrotyrosine, 8-OHdG, OX-LDL, PON-1, FRAP, GSH, or TBARS following the MD intervention. The lack of significant differences may be attributed to the limited number of original studies, which may have reduced the ability to evaluate the true effects. Additionally, after long-term adherence to the MD, compensatory decreases in the activity of antioxidant enzymes, such as SOD, may occur [37]. PON-1 is also influenced by genetic polymorphisms, such as population differences [80–82]. Therefore, more rigorously designed studies are needed to further interpret these findings.

In this study, some pooled results showed significant heterogeneity. This heterogeneity may stem from differences in population characteristics. The included studies comprised healthy individuals, overweight participants, and patients with cardiovascular disease. These populations differed in their baseline oxidative stress status and metabolic profiles. Furthermore, variations in the comparator diets, intervention durations, and biomarker measurement methods across studies may have influenced the pooled results. This heterogeneity may reduce the generalizability of the findings, thereby weakening their credibility.

Limitations

This systematic review and meta-analysis found that the MD enhances antioxidant capacity. However, this study has some limitations. First, the number of original studies available for the meta-analysis was limited by the strict inclusion criteria, which included only studies that evaluated the MD alone. Consequently, the limited number of included studies reduced the statistical power, affected the reliability of the pooled effect estimates, and limited further subgroup analyses. Therefore, we recommend considering these pooled results as preliminary evidence that reflects the overall trends. Additionally, owing to the specific nature of dietary interventions, blinding of participants and investigators was difficult to achieve in most studies, introducing a risk of bias. Finally, geographic differences may have introduced bias, as most studies were conducted in Europe.

Conclusions

The MD significantly increased blood TAC and decreased urine F2-IsoP and blood MDA levels, suggesting that the MD may play a role in enhancing antioxidant capacity and reducing oxidative damage. However, the number of original studies included in this meta-analysis was limited. Therefore, high-quality, large-scale randomized controlled trials with long-term follow-up are required to better elucidate the effects of the MD on oxidative stress levels.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

MD

Mediterranean diet

BMI

Body mass index

SMD

Standardized mean difference

TBARS

Thiobarbituric acid reactive substances

OX-LDL

Oxidized low-density lipoprotein

MDA

Malondialdehyde

F2-IsoP

F2-isoprostanes

8-OHdG

8-hydroxy-2’-deoxyguanosine

TAC

Total antioxidant capacity

FRAP

Ferric reducing antioxidant power

PON-1

Paraoxonase-1

CAT

Catalase

GSH

Glutathione

SOD

Superoxide dismutase

GPX

Glutathione peroxidase

NAFLD

Nonalcoholic fatty liver disease

PRISMA

Preferred reporting items for systematic reviews and Meta-Analyses

PROSPERO

Prospective register of systematic reviews

NF-κB

Nuclear factor kappa-B

Nrf2

Nuclear factor erythroid 2-related factor 2

AMPK

AMP-activated protein kinase

ROS

Reactive oxygen species

Author contributions

H.D. and Y.F., Methodology; Y.F., Y.L., and T.R., Data curation; H.D., B.D., T.R., and W.L., Formal analysis and investigation; T.R. and W.L., Visualization; H.D. and Y.F., Writing-original draft; R.Z. and D.M., Conceptualization; R.Z. and D.M., Writing-review and editing; R.Z., D.M., and Y.F., Project administration; R.Z. and D.M., Validation; H.D., Supervision; R.Z. and D.M., Funding acquisition. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (grant numbers 82402022, 81971433, 82271749); the Clinical research Special Fund of Wu Jieping Medical Foundation (grant number 320.6750.2023-24-4); the National Key R&D Program of China (grant numbers 2021YFC2701700, 2021YFC2701704); the Fundamental Research Funds for the Central University (grant number SCU2023D006); and the Science and Technology of Sichuan Province (grant number 2021YFS0027).

Data availability

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

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.