Effects of resveratrol supplementation on multiple health outcomes: an umbrella review of systematic reviews and meta-analyses of randomized controlled trials

Abstract

Background

Resveratrol has shown health benefits like anti-obesity, antioxidant, and anti-inflammatory effects.

Objectives

To resolve conflicting findings and clarify its impact, our study summarizes resveratrol supplementation’s effects and evaluates its effectiveness across populations.

Methods

We conducted the search through PubMed, Embase, and Web of Science databases from inception to April 30th, 2025, to identify meta-analyses of randomized controlled trials examining the effects of resveratrol supplementation on human health. The quality of included studies was evaluated using the A Measurement Tool to Assess Systematic Reviews (AMSTAR) tool, and the credibility of evidence was assessed using the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) criteria. This study was registered in PROSPERO (CRD 42024554480).

Results

Forty-five systematic reviews with 129 associations between resveratrol supplementation and human health were included, which consists of 68 outcomes. Using the AMSTAR measurement tool, all articles were categorized as moderate-to-high quality. Among 35 significant associations, four were supported by high-certainty evidence according to GRADE, including positive effects of resveratrol supplementation on reducing waist circumference (mean difference [MD] = -0.80 cm, 95% CI = -1.40, -0.20), total cholesterol among overweight adults (MD = -0.19 mmol/L, 95% CI = -0.32, -0.06), diastolic blood pressure (MD = -3.55 mmHg, 95% CI = -5.18, -1.93) and systolic blood pressure (MD = -7.97 mmHg, 95% CI = -10.63, -5.31) among patients with type 2 diabetes. Moreover, moderate-certainty evidence indicated that resveratrol supplementation may promote glucose metabolism, improve renal function, protect endothelial health, enhance working memory, improve hepatic steatosis, lower hormones, and reduce inflammation.

Conclusions

High certainty of evidence suggested that resveratrol supplementation could decline waist circumference, lower blood pressure among patients with type 2 diabetes, and reduce total cholesterol among overweight adults.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12937-026-01319-5.

Introduction

Resveratrol is a polyphenol belonging to the family of stilbenes, predominantly found in grapes and wine. Resveratrol is known for its numerous health benefits, including boosting immunity, slowing aging processes, and mimicking the effect of calorie restriction, resulting in its anti-obesity, antioxidant, and anti-inflammatory activities [1, 2]. In recent years, resveratrol has garnered great attention as a sirtuin-activating compound. Growing clinical trials have indicated its potential to reduce the risk and progression of various chronic diseases, such as type 2 diabetes mellitus (T2DM) [3], cardiovascular disease (CVD) [4], and cognitive disorders [5, 6]. Despite this, evidence regarding its impact on health outcomes has been conflicting. For example, a meta-analysis in 2019 failed to establish an association between resveratrol intervention and glucose levels [7], whereas another meta-analysis in 2021 encompassing 30 randomized controlled trials (RCTs) demonstrated a significant hypoglycemic effect of resveratrol [8]. Herein, it is imperative to objectively and succinctly synthesize these findings with inconsistent results.

Umbrella reviews (URs) facilitate the elimination of conflicting results among studies by summarizing and comparing findings from multiple meta-analyses using a standardized approach [9]. This process aids in identifying consensus or discrepancies in the existing evidence, contributing to a clearer understanding of complex topics. In 2021, an UR identified 11 meta-analyses and provided an overview of the effects of resveratrol supplementation for managing cardiometabolic risk factors in patients with T2DM, metabolic syndrome (MetS), and nonalcoholic fatty liver disease (NAFLD) [10]. Moreover, Molani-Gol et al. conducted an UR incorporating 19 meta-analyses with 81 unique RCTs and suggested the beneficial effects of resveratrol supplementation on reducing general and central obesity, as well as decreasing some inflammatory markers [11]. These URs indicated that resveratrol’s effectiveness may vary across populations, which has not been compared in previous studies. For example, positive effects were observed among the participants with no restriction on baseline health status and patients with NAFLD. In contrast, no effect was demonstrated in reducing body mass index (BMI) in the interventions with resveratrol among patients with T2DM or MetS. Moreover, these two URs focused on specific areas, leaving some health effects of resveratrol inconclusive, such as improving renal and cognitive function, along with several conflicting findings requiring further resolution. For example, regarding the effect of resveratrol on luteinizing hormone (LH) among patients with polycystic ovary syndrome (PCOS), three meta-analyses have been conducted, which involved the same three RCTs [12–14]. However, only Balieiro et al. suggested that resveratrol supplementation was associated with a reduction in LH [13], while no significant association was found in the other two meta-analyses [12, 14], which might be attributed to miscalculations. Additionally, several high-quality meta-analyses have been published in the past several years that could update evidence [15, 16]. Recognizing these limitations, reviewing the existing systematic reviews, and updating the conclusion encompassing all health-related outcomes related to resveratrol intervention is crucial.

Therefore, we conducted the current UR to comprehensively summarize the health effect of resveratrol supplementation based on meta-analyses of RCTs and to resolve conflicting findings. Our primary aim was to provide an overview of the strength of evidence, the extent of potential biases, and the validity of the claimed effects of resveratrol on human health. Furthermore, the secondary aim was to compare the effect of resveratrol intervention among different populations and identify the population that would benefit the most from resveratrol supplementation, thus supporting evidence-based clinical decision-making.

Method

Protocol registration

The Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines were followed while conducting the current UR (Supplementary Table 1) [17]. The PRISMA statement consists of a 27-item checklist and a four-phase flow diagram, which helps to ensure transparent and comprehensive reporting [17]. Moreover, the protocol for the UR was registered with the International Prospective Register of Systematic Reviews [PROSPERO (https://www.crd.york.ac.uk/PROSPERO/); registration number CRD 42024554480] to avoid duplication and reduce the possibility of reporting bias.

Search strategy

We performed a comprehensive search in PubMed, Embase, and Web of Science databases from inception until April 30th, 2025, for eligible systematic reviews with meta-analyses of RCTs examining the effects of resveratrol supplementation on multiple health outcomes. Our search strategy employed a combination of keywords related to resveratrol, health outcomes, and meta-analysis. The complete search strategy is available in Supplementary Table 2. We also manually reviewed references from eligible systematic reviews and meta-analyses. Two investigators (J-NS and R-Y) retrieved potentially eligible articles independently, and any discrepancies were resolved by a third investigator (Q-JW).

Selection criteria

Two investigators (J-NS and R-Y) independently screened titles and abstracts for relevance and assessed the full texts of potentially eligible articles. Any discrepancies were resolved by discussion with a third investigator (Q-JW). Studies in the English language in accordance with the following Population, Intervention, Comparator, Outcome, and Study design (PICOS) criteria were included:

1. Population: adults of any ethnicity in any country or setting;

2. Intervention: resveratrol alone, resveratrol along with other standard treatment (e.g., physical rehabilitation, or hypoglycemic agents), or higher dose of resveratrol;

3. Comparator: placebo alone (e.g., maltodextrin), placebo along with other standard treatment (e.g., physical rehabilitation or hypoglycemic agents), or lower dose of resveratrol

4. Outcome: any health outcomes or indicators, such as anthropometric measures (e.g., BMI), lipid profile outcomes (e.g., triacylglycerols), glycemic profile outcomes (e.g., insulin level), inflammation biomarker (e.g., C-reactive protein), and biomarkers of liver health (e.g., alkaline phosphatase), etc.; and

5. Study design: systematic reviews and meta-analyses of RCTs.

The exclusion criteria were as follows: (1) systematic reviews without quantitative analysis; (2) studies that focused on observational studies, laboratory studies, or animal studies; (3) studies that did not report comprehensive data (such as 95% confidence intervals [CIs], effect sizes, and the number of participants); or (4) meta-analyses including fewer than three RCTs.

When selecting among multiple meta-analyses addressing the same scientific question, the one with the largest number of studies was retained for the primary analysis [18]. In instances with similar datasets, we preferred the meta-analysis with the highest methodological quality. Furthermore, each analysis was assessed separately if an article performed different meta-analyses on more than one outcome or among various populations.

Data extraction

Two trained investigators (R-Y and L-F) worked in pairs to independently collect the information from each eligible study. All disagreements were resolved through consultation with the senior investigator (Q-JW). Data authenticity was verified for each study included in the review. We abstracted information on the first author, publication year, journal, exposure, number of included RCTs, heterogeneity I², P-values, pooled effect (mean difference [MD] or standardized mean difference [SMD]), and corresponding 95% CIs, and outcomes for each eligible systematic review and meta-analysis. For primary studies in each eligible meta-analysis, we further extracted the first author, publication year, baseline health status of participants, number of cases, number of controls, number of total participants, type of intervention (type, duration, and doses), the treatment in control groups, and study-specific risk estimates and 95% CIs.

Assessment of methodological and credibility quality of evidence

The validated A Measurement Tool to Assess Systematic Reviews (AMSTAR) instrument was used to assess methodological quality and assign an overall rating for each included systematic review and meta-analysis [19]. AMSTAR is a reliable and valid measurement tool containing 11 items to assess the methodological quality of systematic reviews. The AMSTAR score was categorized as high (8 − 11), moderate (4 − 7), and low (0 − 3) quality [19].

Moreover, the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) was applied to evaluate the credibility of evidence as high, moderate, low, or very low quality [20]. RCTs are initiated as high-certainty evidence that could be downgraded based on prespecified criteria. The criteria assigned to downgrade evidence include the risk of bias (assessed by the Cochrane Risk of Bias instrument, inconsistency, indirectness, imprecision, and compelling evidence of publication bias. In cases where the systematic review lacked an assessment of bias risk, we utilized the Cochrane tool to assess the risk of bias in included RCTs.

Four investigators (J-NS, L-F, R-Y, and Y-JL) worked in pairs to independently evaluate the methodological quality of eligible systematic reviews based on AMSTAR, and the credibility of evidence for each association according to GRADE. Subsequently, any disagreements in ratings were resolved by consensus after further discussion with a senior investigator (Q-JW).

Sensitivity analyses and subgroup analyses

We conducted several sensitivity analyses for associations supported by moderate-to-high certainty to evaluate the robustness and reliability of pooled results in the primary analysis. RCTs with small sample sizes (< 25th percentile) and those with a high risk of bias were excluded from identified associations [21, 22]. Moreover, subgroup analysis was conducted based on the intervention duration and the dose of resveratrol supplementation to identify any sources of heterogeneity [11]. The certainty of evidence for each subgroup was also evaluated using the GRADE criteria. All statistical analyses were performed in STATA version 17.

Result

Literature review

A PRISMA flow chart depicting the search process and study selection is shown in Fig. 1. After searching three databases (PubMed, Embase, and Web of Science) and removing duplicate articles, 1146 articles were initially identified. We removed 869 articles based on reviewing the title and abstract. Of 248 articles reviewed in full-text, 45 systematic reviews and meta-analyses of RCTs met our inclusion criteria and were included in this UR. Two investigators had a highly consistent agreement on the selection of studies, with kappa value of 0.963. Details of excluded meta-analyses are reported in Supplementary Table 3.

Fig. 1.

Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) flow diagram of study screening and selection

Characteristics of the included meta-analyses

The 45 eligible articles encompassed 129 associations, investigating 68 outcomes of human health and resveratrol intervention (Table 1). These meta-analyses were published from 2013 to 2025. The median number of included primary studies in each association was 5 (range 3-30), and the median number of participants was 296 (range 66-1747). The present UR population comprised healthy adults, overweight participants, and individuals with comorbidities such as T2DM, MetS, NAFLD, and PCOS. Of the 129 associations, a wide range of outcomes was evaluated: lipid profile (n = 23, 17.9%), inflammation and oxidative stress (n = 20, 15.5%), anthropometric measure outcomes (n = 15, 11.6%), endothelial function and blood pressure (n = 15, 11.6%), glycemic profile (n = 15, 11.6%), liver health (n = 12, 9.3%), cognitive function (n = 7, 5.4%), renal function (n = 7, 5.4%), bone health (n = 5, 3.9%), hormones (n = 4, 3.1%), Sirtuin 1 (n = 3, 2.3%), mental health (n = 2, 1.6%), and physical activity (n = 1, 0.8%).

Table 1.

Description of 129 associations between resveratrol (RV) intervention and multiple health outcomes

Outcomes	Population	Author, year, reference	No. of RCTs (sample size)	Metric	Dose of RV intervention (mg/d)	Duration of RV intervention (weeks)	Comparator
Anthropometric measures
Body mass index	No limitation	Tabrizi, 2020 [23]	27 (1277)	SMD	100–1500	4–24	Placebo, or with hypoglycemic agents
Body mass index	Patients with NAFLD	Rafiee, 2020 [24]	6 (266)	SMD	300–3000	8–24	Placebo
Body weight	No limitation	Tabrizi, 2020 [23]	25 (1124)	SMD	100–1500	4–26	Placebo
Body weight	Patients with MetS	Asgary, 2019 [7]	8 (317)	SMD	100–3000	3–16	Placebo
Body weight	Patients with NAFLD	Rafiee, 2020 [24]	6 (266)	SMD	300–3000	8–24	Placebo
Fat mass	No limitation	Mousavi, 2019 [25]	8 (458)	MD	40-1000	4–24	Placebo
Fat percentage	No limitation	Zhou, 2022 [26]	5 (202)	SMD	50-1000	4–26	Placebo
Intra-hepatic fat content	No limitation	Soltani, 2024 [16]	6 (276)	MD	8-3000	4–48	Placebo
Lean mass	No limitation	Tabrizi, 2020 [23]	5 (241)	SMD	100–1000	4–16	Placebo, or placebo with caloric restriction
Liver fat	Patients with NAFLD	Kilchoer, 2020 [27]	4 (170)	MD	150–3000	8–26	Placebo
Waist circumference	No limitation	Mousavi, 2019 [25]	8 (663)	MD	40-1000	4–24	Placebo
Waist circumference	Patients with MetS	Asgary, 2019 [7]	5 (171)	SMD	100–1000	9–12	Placebo
Waist circumference	Patients with NAFLD	Elgebaly, 2017 [28]	4 (110)	MD	300–3000	8–26	Placebo
Waist circumference	Patients with T2DM	Gu, 2022 [29]	5 (459)	MD	40-1000	8–24	Placebo
Waist to hip ratio	Patients with NAFLD	Rafiee, 2020 [24]	4 (196)	SMD	300–3000	8–24	Placebo
Bone biomarkers
Serum carboxy-terminal peptide	No limitation	Li, 2021 [30]	3 (218)	MD	100–3000	16–48	Placebo
Serum osteocalcin	No limitation	Li, 2021 [30]	3 (218)	MD	100–3000	16–48	Placebo
Bone mineral density
Lumbar spine	No limitation	Li, 2021 [30]	3 (371)	MD	100–3000	16–48	Placebo
Total hip	No limitation	Li, 2021 [30]	3 (373)	MD	100–3000	16–48	Placebo
Whole body	No limitation	Li, 2021 [30]	3 (373)	MD	100–3000	16–48	Placebo
Cognitive function
Attention speed	No limitation	Khorshidi, 2021 [31]	4 (184)	SMD	72-1000	12–26	Placebo alone, or with quercetin
Delayed recall	No limitation	Khorshidi, 2021 [31]	4 (185)	SMD	72-1000	14–26	Placebo
Delayed recognition	No limitation	Marx, 2018 [32]	4 (166)	SMD	150–250	14–26	Placebo
Episodic memory	No limitation	de Vries, 2021 [33]	3 (611)	SMD	75–200	6–26	Placebo
Executive functions	No limitation	Farag, 2023 [34]	4 (192)	SMD	75-1000	12–26	Placebo
Immediate memory	No limitation	Khorshidi, 2021 [31]	5 (217)	SMD	72-1000	12–26	Placebo
Working memory	No limitation	de Vries, 2021 [33]	4 (424)	SMD	75–200	6–26	Placebo, or with quercetin
Endothelial function and blood pressure
Diastolic blood pressure	No limitation	Akbari, 2019 [35]	27 (1591)	SMD	100–1502	1–50	Placebo alone, with physical rehabilitation, or with standard treatment for MetS
Diastolic blood pressure	Healthy participants	Weaver, 2020 [36]	3 (253)	MD	75-2457	2–12	Placebo
Diastolic blood pressure	Overweight participants	Liu, 2014 [37]	4 (171)	MD	16-1000	4–48	Placebo
Diastolic blood pressure	Patients with NAFLD	Huang, 2024 [38]	5 (261)	MD	150–3000	8–24	Placebo
Diastolic blood pressure	Patients with T2DM	Gu, 2022 [29]	9 (435)	MD	8-1000	1.5–24	Placebo, or with standard treatment for T2DM
Flow-mediated dilatation	No limitation	Mohammadipoor, 2022 [39]	8 (603)	MD	30–500	1–24	Placebo
Flow-mediated dilatation	Patients with MetS	Akbari, 2019 [35]	4 (234)	SMD	75–300	1–12	Placebo
Mean arterial pressure	No limitation	Fogacci, 2019 [40]	17 (746)	MD	8-3000	1.5–48	Placebo, or with hypoglycemic agents
Pulse pressure	No limitation	Fogacci, 2019 [40]	17 (746)	MD	8-3000	1.5–48	Placebo, or with hypoglycemic agents
Systolic blood pressure	No limitation	Akbari, 2019 [35]	27 (1591)	SMD	100–1500	1–48	Placebo alone, with physical rehabilitation, or with standard treatment for MetS
Systolic blood pressure	Healthy participants	Weaver, 2020 [36]	3 (253)	MD	75-2457	2–12	Placebo
Systolic blood pressure	Patients with MetS	Asgary, 2019 [7]	3 (143)	SMD	150–1000	4–12	Placebo
Systolic blood pressure	Overweight participants	Liu, 2014 [37]	4 (171)	MD	16-1000	4–48	Placebo
Systolic blood pressure	Patients with NAFLD	Huang, 2024 [38]	5 (261)	MD	150–3000	8–24	Placebo
Systolic blood pressure	Patients with T2DM	Gu, 2022 [29]	9 (435)	MD	8-1000	1.5–24	Placebo, or usual medical care
Glycemic profile
Fasting blood glucose	No limitation	Delpino, 2022 [3]	23 (1011)	SMD	120–3000	4–28	Placebo
Fasting blood glucose	Patients with T2DM	Gu, 2022 [29]	19 (1151)	MD	8-1500	4–48	Placebo, or usual medical care
Fasting blood glucose	Patients with NAFLD	Huang, 2024 [38]	5 (296)	SMD	150–3000	8–12	Placebo
Glucose level	No limitation	García-Martínez, 2021 [8]	30 (1651)	MD	40-1000	2–16	Placebo
Glucose level	Patients with MetS	Asgary, 2019 [7]	5 (265)	SMD	100–600	4–12	Placebo
HbA1c	No limitation	Delpino, 2022 [3]	12 (596)	SMD	150–3000	4–26	Placebo, or placebo with losartan
HbA1c	Overweight participants	Huang, 2016 [41]	5 (196)	MD	100–1000	4–26	Placebo
HbA1c	Patients with T2DM	Beatriz Isabel, 2022 [42]	12 (600)	MD	150–3000	4–24	Placebo
HOMA-IR	No limitation	Delpino, 2022 [3]	13 (612)	SMD	10-1000	4–16	Placebo
HOMA-IR	Patients with NAFLD	Huang, 2024 [38]	6 (321)	MD	150–3000	8–24	Placebo
HOMA-IR	Patients with NAFLD	Zhang, 2016 [43]	4 (156)	MD	150–3000	8–26	Placebo
HOMA-IR	Overweight participants	Huang, 2016 [41]	8 (286)	MD	100–3000	2–48	Placebo
Insulin level	Patients with NAFLD	Zhang, 2016 [43]	4 (156)	MD	150–3000	8–26	Placebo
Insulin level	No limitation	García-Martínez, 2021 [8]	24 (1293)	MD	40-1000	2–16	Placebo
Insulin level	Overweight participants	Zahoor, 2024 [44]	5 (267)	SMD	150–500	4–24	Placebo
Insulin level	Patients with T2DM	Nyambuya, 2020 [45]	3 (409)	SMD	500–3000	2–12	Placebo
Hormones
Follicle-stimulating hormone	Patients with PCOS	Fadlalmola, 2023 [12]	3 (169)	SMD	800–1500	5.7–12	Placebo
Luteinizing hormone	Patients with PCOS	Balieiro, 2024 [13]	3 (169)	MD	800–1500	5.7–12	Placebo
Prolactin	Patients with PCOS	Fadlalmola, 2023 [12]	3 (169)	SMD	800–1500	5.7–12	Placebo
Testosterone	Patients with PCOS	Fadlalmola, 2023 [12]	3 (169)	SMD	800–1500	5.7–12	Placebo
Inflammation and oxidative stress
Catalase	No limitation	Koushki, 2019 [46]	4 (126)	SMD	100–3000	4–17	Placebo
C-reactive protein	No limitation	Gorabi, 2021 [47]	15 (683)	MD	8-1000	4–48	Placebo, or with usual medical care
C-reactive protein	Patients with CVD	Teimouri, 2022 [48]	4 (200)	MD	8-100	8–51	Placebo
C-reactive protein	Patients with MetS or related disease	Tabrizi, 2018 [49]	19 (892)	SMD	8-3000	4–24	Placebo
C-reactive protein	Patients with PCOS	Fadlalmola, 2023 [12]	3 (148)	MD	800–1000	6–12	Placebo
C-reactive protein	Patients with T2DM	Zhu, 2025 [50]	5 (563)	SMD	40-1000	8–24	Placebo
Glutathione peroxidase	No limitation	Koushki, 2019 [46]	4 (120)	SMD	100–3000	4–17	Placebo
Interleukin-6	No limitation	Omraninava, 2021 [51]	17 (851)	MD	8-3000	4–48	Placebo
Interleukin-6	Patients with CVD	Teimouri, 2022 [48]	3 (140)	MD	8-100	8–51	Placebo
Interleukin-6	Patients with T2DM	Zhu, 2025 [50]	4 (435)	SMD	40–800	4–24	Placebo
Interleukin-6	Participants with CAD risk	Damay, 2024 [52]	5 (218)	MD	8.1–800	4–48	Placebo, or with hypoglycemic agents, or with antihypertensive agents
Malondialdehyde	No limitation	Omidian, 2020 [53]	4 (222)	MD	500–800	6–12	Placebo
Superoxide dismutase	No limitation	Koushki, 2019 [46]	8 (325)	SMD	100–3000	4–17	Placebo
Glutathione peroxidase	No limitation	Koushki, 2019 [46]	4 (120)	SMD	100–3000	4–17	Placebo
TNF-α	No limitation	Omraninava, 2021 [51]	17 (812)	MD	8-3000	4–48	Placebo
TNF-α	Patients with CVD	Teimouri, 2022 [48]	5 (255)	MD	8-100	8–51	Placebo
TNF-α	Patients with NAFLD	Zhang, 2016 [43]	4 (156)	MD	150–3000	8–26	Placebo
TNF-α	Participants with CAD risk	Damay, 2024 [52]	4 (189)	MD	8.1–800	4–48	Placebo, or with hypoglycemic agents, or with antihypertensive agents
TNF-α	Patients with confirmed CAD	Damay, 2024 [52]	3 (247)	MD	8.1–100	8–48	Placebo, or with hypoglycemic agents, or with antihypertensive agents
TNF-α	Patients with T2DM	Zhu, 2025 [50]	3 (198)	SMD	200–800	4–24	Placebo
Total antioxidant capacity	No limitation	Koushki, 2019 [46]	8 (400)	SMD	100–3000	4–17	Placebo
Lipid profile
Adiponectin	No limitation	Tabrizi, 2020 [23]	10 (563)	SMD	8-1000	4–48	Placebo
Apoprotein-A	Patients with NAFLD	Rafiee, 2020 [24]	3 (160)	SMD	300–600	12	Placebo
Free fatty acids	No limitation	Sahebkar, 2013 [54]	3 (66)	MD	40-1500	4–6	Placebo
High density lipoprotein	No limitation	Akbari, 2020 [55]	26 (1431)	MD	20-3000	4–48	Placebo
High density lipoprotein	Overweight participants	Huang, 2016 [41]	13 (538)	MD	8-3000	4–26	Placebo
High density lipoprotein	Patients with MetS	Asgary, 2019 [7]	5 (265)	SMD	100–600	4–12	Placebo
High density lipoprotein	Patients with NAFLD	Huang, 2024 [38]	6 (321)	SMD	150–3000	8–24	Placebo
High density lipoprotein	Patients with T2DM	Gu, 2022 [29]	14 (837)	MD	8-3000	1.5–48	Placebo
Low density lipoprotein	No limitation	Akbari, 2020 [55]	24 (1291)	MD	20-3000	4–48	Placebo
Low density lipoprotein	Overweight participants	Huang, 2016 [41]	12 (493)	MD	8-3000	4–26	Placebo
Low density lipoprotein	Patients with NAFLD	Huang, 2024 [38]	6 (271)	SMD	150–3000	8–24	Placebo
Low density lipoprotein	Patients with T2DM	Zhang, 2021 [56]	10 (661)	MD	50-1000	4–24	Placebo
Leptin	No limitation	Zhou, 2022 [26]	7 (271)	SMD	50–250	4–26	Placebo
Total cholesterol	No limitation	Akbari, 2020 [55]	25 (1435)	MD	20-3000	4–48	Placebo
Total cholesterol	Overweight participants	Huang, 2016 [41]	12 (524)	MD	8-3000	4–26	Placebo
Total cholesterol	Patients with MetS	Asgary, 2019 [7]	4 (231)	SMD	100–600	4–12	Placebo
Total cholesterol	Patients with NAFLD	Huang, 2024 [38]	5 (295)	SMD	150–3000	8–12	Placebo
Total cholesterol	Patients with T2DM	Zhang, 2021 [56]	10 (730)	MD	50-1000	4–24	Placebo
Triglyceride	No limitation	Akbari, 2020 [55]	28 (1543)	MD	20-3000	4–48	Placebo, or with usual medical care
Triglyceride	Patients with MetS	Asgary, 2019 [7]	5 (226)	SMD	100–600	4–12	Placebo
Triglyceride	Overweight participants	Huang, 2016 [41]	14 (544)	MD	8-3000	4–26	Placebo
Triglyceride	Patients with NAFLD	Huang, 2024 [38]	6 (321)	SMD	150–3000	8–24	Placebo
Triglyceride	Patients with T2DM	Gu, 2022 [29]	15 (976)	MD	8-3000	1.5–48	Placebo
Liver health
Alkaline phosphatase	No limitation	Soltani, 2023 [16]	15 (926)	MD	8-3000	4–48	Placebo
Alkaline phosphatase	Patients with T2DM	Li, 2021 [30]	3 (270)	MD	8-1000	6–48	Placebo
Alanine aminotransferase	No limitation	Soltani, 2023 [16]	30 (1747)	MD	8-3000	4–48	Placebo
Alanine aminotransferase	Patients with NAFLD	Huang, 2024 [38]	6 (321)	MD	150–3000	8–24	Placebo
Alanine aminotransferase	Overweight participants	Zahoor, 2024 [44]	3 (87)	SMD	150–500	4–24	Placebo
Aspartate aminotransferase	No limitation	Soltani, 2023 [16]	30 (1607)	MD	8-3000	4–48	Placebo
Aspartate aminotransferase	Patients with NAFLD	Huang, 2024 [38]	6 (321)	MD	150–3000	8–24	Placebo
Improvement in hepatic steatosis	Patients with NAFLD	Huang, 2024 [38]	3 (170)	OR	300–600	8–12	Placebo
Total bilirubin	No limitation	Soltani, 2023 [16]	15 (679)	MD	8-3000	4–48	Placebo
Total bilirubin	Patients with NAFLD	Elgebaly, 2017 [28]	4 (129)	MD	300–3000	8–24	Placebo
γ-glutamyl transpeptidase	No limitation	Soltani, 2023 [16]	18 (1027)	MD	8-3000	4–48	Placebo
γ-glutamyl transpeptidase	Patients with NAFLD	Zhang, 2016 [43]	3 (136)	MD	150–500	12–25	Placebo
Mental health
Depression	No limitation	Khorshidi, 2021 [31]	4 (180)	SMD	72-1000	4–24	Placebo
Negative mood	No limitation	Marx, 2018 [32]	3 (928)	SMD	150–250	0.5–14	Placebo
Physical activity
Peak oxygen uptake	Healthy participants	Srilacorn, 2018 [57]	5 (139)	MD	125–500	4	Placebo
Renal function
Albumin	No limitation	Abdollahi, 2023 [58]	10 (418)	MD	8-1000	4–48	Placebo
Blood urea nitrogen	No limitation	Abdollahi, 2023 [58]	17 (765)	MD	8-1000	4–48	Placebo
Creatinine	No limitation	Abdollahi, 2023 [58]	24 (1022)	MD	8-1000	4–48	Placebo
Creatinine	Patients with T2DM	Nyambuya, 2020 [45]	5 (209)	SMD	250–3000	4–12	Placebo
Glomerular filtration rate	No limitation	Abdollahi, 2023 [58]	3 (127)	MD	8-1000	4–48	Placebo
Urinary protein	No limitation	Abdollahi, 2023 [58]	7 (302)	MD	8-1000	4–48	Placebo
Uric acid	No limitation	Abdollahi, 2023 [58]	4 (274)	MD	8-1000	4–48	Placebo
Sirtuin 1
Sirtuin 1 gene expression level	No limitation	Mansouri, 2025 [59]	4 (344)	SMD	75-1000	4–16	Placebo, or with low-calorie diet
Sirtuin 1 protein expression level	No limitation	Mansouri, 2025 [59]	5 (351)	SMD	40-3000	4–24	Placebo
Sirtuin 1 serum level	No limitation	Mansouri, 2025 [59]	3 (293)	MD	500–1000	4–24	Placebo, or with low-calorie diet

CAD coronary artery disease, CVD cardiovascular disease, HbA1c hemoglobin A1c, HOMA-IR homeostasis model assessment of insulin resistance, MetS metabolic syndrome, MD mean difference, NAFLD non-alcoholic fatty liver disease, PCOS polycystic ovary syndrome, RCT randomized controlled trial, SMD standard mean difference, TNF-α tumor necrosis factor-α, T2DM type 2 diabetes mellitus

Methodological quality of meta-analyses

According to the results of the quality assessment by the AMSTAR questionnaire (Fig. 2 and Supplementary Table 4), 26 (57.8%) studies were rated as high confidence, and 19 (42.2%) were rated as moderate confidence. The median AMSTAR score was 8 (range 5–10). The AMSTAR assessment was downgraded mainly because of the absence of a comprehensive list outlining the included and excluded studies.

Fig. 2.

Methodological quality assessment of the included articles with AMSTAR. AMSTAR, A Measurement Tool to Assess Systematic Reviews

Summary effect size and certainty of evidence

Thirty-five associations (27.1%) were of nominally statistical significance (P < 0.05) among the 129 associations, which were listed in Fig. 3 and Supplementary Table 5. Of these, resveratrol intervention has a significant effect on nine outcomes, including glycemic profile (n = 7), inflammation and oxidative stress (n = 7), anthropometric measure outcomes (n = 5), endothelial function and blood pressure (n = 5), lipid profile (n = 3), renal function (n = 3), cognitive function (n = 2), hormone levels (n = 2), and liver health (n = 1) (Fig. 3).

Fig. 3.

The distribution of the certainty of evidence in statistically significant associations

According to the GRADE, four significant associations were supported by high certainty of evidence, 23 by moderate certainty of evidence, 5 rated as low certainty of evidence, and 3 as very low certainty of evidence (Fig. 3). Among the 23 moderate-certainty associations, 13 were downgraded due to high heterogeneity (I² > 50%), nine were downgraded for imprecision caused by small numbers of participants, and one was downgraded due to indirection (Supplementary Table 5).

Figure 4 shows the strength and credibility of the 27 significant associations supported by moderate-to-high certainty of evidence, including 4 positive and 23 inverse associations between resveratrol and multiple health outcomes. Rated as high certainty, resveratrol has positive effects on reducing the level of total cholesterol among overweight adults (MD = -0.19 mmol/L, 95% CI = -0.32, -0.06), decreasing the waist circumference (MD = -0.80 cm, 95% CI = -1.40, -0.20), and hypotensive effects among patients with T2DM (MD = -3.55 mmHg, 95% CI = -5.18, -1.93 for diastolic blood pressure; and MD = -7.97 mmHg, 95% CI = -10.63, -5.31 for systolic blood pressure) (Fig. 4).

Fig. 4.

The strength of significant associations supported by moderate-to-high quality according to the GRADE criteria. The associations displayed in this figure were selected based on statistical significance and certainty of evidence from the total 129 associations between resveratrol supplementation and multiple health outcomes. AMSTAR, A Measurement Tool to Assess Systematic Reviews; CAD, coronary artery disease; CI, confidence interval; GRADE, Grading of Recommendations Assessment, Development, and Evaluation; MD, mean difference; MetS, metabolic syndrome; NAFLD, Non-alcoholic fatty liver disease; SMD, standard mean difference; RCT, randomized controlled trial

Evidence rated as moderate certainty and with low heterogeneity suggested that resveratrol intervention would enhance working memory (SMD = 0.20, 95% CI = 0.03, 0.36), increase the glomerular filtration rate (MD = 7.58 mL/min/1.73 m², 95% CI = 5.25, 9.92), reduce some hormone levels among PCOS patients (SMD = -0.40 ng/mL, 95% CI = -0.71, -0.10 for testosterone; and MD = -2.32 IU/mL, 95% CI = -4.12, -0.53 for LH), decrease the plasma level of inflammatory factors among patients with CVD (MD = -0.55 mg/L, 95% CI = -1.04, -0.06 for C-reactive protein; and MD = -0.63 pg/mL, 95% CI = -1.13, -0.13 for tumor necrosis factor-α [TNF-α]), and reduce TNF-α level among patients with coronary artery disease (MD = -0.47, 95% CI = -0.73, -0.20) (Fig. 4). Moreover, among patients with NAFLD, moderate certainty of evidence with low heterogeneity supported that resveratrol intervention would lower glucose level (MD = -0.29, 95% CI = -0.52, -0.06), decrease diastolic blood pressure (MD = -0.32, 95% CI = -0.46, -0.19) and improve hepatic steatosis (SMD = 0.64, 95% CI = 0.10, 1.18) (Fig. 4).

In addition, high certainty of evidence supported that resveratrol intake did not significantly influence mean arterial pressure, pulse pressure, alkaline phosphatase, total bilirubin, and albumin among participants with no restriction. Among patients with T2DM, increased resveratrol intake did not result in significant changes in patients’ waist circumference or HDL levels. Among overweight patients, resveratrol also showed no significant effect on blood lipid regulation, including HDL and TG levels (Supplementary Table 6).

Sensitivity analyses

We performed several sensitivity analyses for evidence rated over moderate certainty using the GRADE system (Supplementary Tables 7 and 8). After excluding studies with a small sample size (< 25th percentile) from the 66 eligible associations rated as over moderate certainty, thirty-one associations remained the same as the primary analysis (Supplementary Table 7).

Forty-five meta-analyses were eligible for the sensitivity analysis, excluding studies with a high risk of bias. As a result, twenty-seven associations remained over moderate certainty, fifteen were downgraded as low certainty, and three decreased to very low certainty (Supplementary Table 8).

Subgroup analyses

The results of subgroup analyses based on dose (< 500 vs. ≥ 500 mg/d) and duration (< 12 vs. ≥ 12 weeks) of resveratrol intervention were shown in Supplementary Tables 9 and 10. According to subgroup analysis based on doses, moderate certainty of evidence supported that intervention with lower dose was significantly associated with declined waist circumference, lower body weight, lower blood urea nitrogen (BUN), and lower triglyceride levels. Moreover, resveratrol dose < 500 mg/d significantly decreased the level of low-density lipoprotein cholesterol among overweight adults, decreased triglyceride levels among patients with T2DM, and reduced CRP and TNF-α levels among patients with CVD. Among patients with NAFLD, moderate certainty of evidence supported that the intervention of resveratrol with higher dose slightly increased the level of lower triglyceride levels (Supplementary Table 9).

When stratified by intervention duration, high certainty of evidence showed that longer term of resveratrol supplementation would increase the level of alkaline phosphatase. Additionally, resveratrol intervention ≥ 12 weeks was significantly associated with decreased TNF-α among CVD patients, lower BUN levels, and declined waist circumference. Resveratrol intervention < 12 weeks exhibited stronger effects on lowering blood pressure among T2DM patients, compared with the main analysis. Both durations of intervention contributed to lowering insulin levels, but the certainty of evidence varied between the two subgroups (Supplementary Table 10).

Discussion

Main Findings

This UR first summarized and evaluated the evidence of resveratrol intervention with diverse health outcomes. According to the GRADE criteria, our study confirmed the beneficial effects of resveratrol treatment on declining waist circumference, reducing the total cholesterol level among overweight adults, and hypotensive effects on both diastolic and systolic blood pressure among patients with T2DM, which were supported by high certainty of evidence. Moreover, we identified moderate-certainty evidence for the positive effects of resveratrol supplementation in promoting glucose metabolism, improving renal function, protecting endothelial health, enhancing working memory, lowering hormones, improving hepatic steatosis, and reducing inflammation.

In the current study, high-to-moderate certainty of evidence supported the effect of resveratrol intervention in combating obesity, especially in reducing central adiposity. The UR conducted by Molani‑Gol et al. in 2023 found that the supplementation of resveratrol led to a reduction in waist circumference, and both doses of resveratrol (≥ 500 and < 500 mg daily) contributed to the reduction [11]. As a theory put forth to explain the effect of resveratrol against obesity, recent studies have paid much attention to the fact that resveratrol may act as a fat browning activator, involved in the secretion of many myokines and adipokines, which helps in maintaining a healthy weight against obesity [60]. Pertinent to metabolic syndrome, glucose metabolism and insulin sensitivity also play a crucial role in combating obesity. Based on the current findings, resveratrol was related to a significant decrease in glucose (− 5.24 mg/dL, P = 0.002) and insulin levels (− 1.23 mIU/L, P = 0.0003) compared with the control group. However, our findings indicated that resveratrol had no impact on glucose levels or insulin levels among patients with T2DM or NAFLD. In contrast, Zeraattalab-Motlagh et al. summarized the results of eight RCTs including 314 patients with T2DM, and indicated that resveratrol supplementation could decrease fasting glucose levels in patients with T2DM [10]. Nevertheless, the number of participants in their study was relatively small. Additionally, they utilized 1.6 mmol/L as minimal clinically important differences for fasting glucose level and acknowledged that the effect was too small to be considered clinically meaningful (MD = -0.33 mmol/L, 95% CI = -0.57, -0.09) [10]. It is worth noting that the impact of resveratrol on glucose metabolism and insulin sensitivity might differ among individuals with different health status [61, 62]. In 2024, Park et al. carried out an experiment among 8 lean and 8 severely obese participants to determine whether obesity status would affect the response to resveratrol in human skeletal muscle and the efficacy of resveratrol in improving insulin action [61]. As a result, resveratrol improved insulin-mediated signal transduction (such as insulin-stimulated phosphorylation of IRS1) in the myotubes of lean individuals, but not in the myotubes of individuals with severe obesity [61]. In addition, an experiment in 2022 utilized perfused livers of healthy and arthritic rats to investigate the effects of resveratrol on glucose metabolism [62]. The liver of arthritic rats was more sensitive to resveratrol, particularly concerning gluconeogenesis, and the reason could be the consequence of a more inflamed and oxidized state of the organ [62]. Considering the high heterogeneity among the results of existing studies [3, 8, 29], further trials with more refined subgroup analyses are warranted to address the unanswered questions and controversies regarding the true benefits of resveratrol for glucose metabolism and insulin sensitivity.

The available evidence rated as moderate certainty from our UR signifies that resveratrol demonstrated efficacy in reducing testosterone and LH among patients with PCOS, a disease characterized by hyperandrogenic symptoms and ovarian dysfunction [63]. Aligning with our findings, a systematic review of eight animal studies and three clinical studies underlined the potential effects of resveratrol on ameliorated ovarian volume, high-quality oocyte rate and embryo rate, as well as concentrations of androgens and gonadotropins in patients with PCOS [64]. In recent years, growing evidence has intensively focused on the benefits of resveratrol on ovarian function. A meta-analysis of 25 preclinical studies in 2024 highlighted that treatment with natural products, including resveratrol, was associated with increased primordial follicles and levels of anti-Müllerian hormone in premature ovarian failure models [65]. Furthermore, resveratrol exhibits the potential to influence the behavior of ovarian cancer chorioallantois membrane tumors, suggesting its consideration as a complementary treatment option for women with incurable ovarian cancer [66, 67]. Collectively, these findings suggested that resveratrol could serve as a viable complementary medicine adjunct for the prevention and treatment of PCOS by improving ovarian function. However, due to the limited number of RCTs evaluating the function of resveratrol on ovarian function and hormones, comprehensive subgroup analyses according to the dose and duration of resveratrol intervention were not feasible. Additional robust clinical trials and in-vivo laboratory studies are warranted in the future to facilitate the broader application of resveratrol.

Evidence from this UR indicated that resveratrol intervention has a protective impact on cardiovascular health. Resveratrol effectively enhances endothelial health by improving flow-mediated dilation response (1.4% increase compared with the controlled group), supported by moderate certainty of evidence. The results aligned with the previous review that assessed the benefits of resveratrol in clinical trials, and its potential to improve vascular perfusion following endothelial vasodilator function through improved flow-mediated dilation [68]. The underlying mechanism behind the effect of resveratrol on cardiovascular health involves multiple metabolic pathways. Resveratrol has been demonstrated to reduce wall-stiffening components in resistant arteries, such as collagen and elastin [69]. Also, the impact of resveratrol on endothelial activity, primarily through nitric oxide and sirtuin pathways, has been previously substantiated [70]. Furthermore, resveratrol was capable of modulating the expression of vascular endothelial growth factor and the formation of the vascular network in a biphasic pattern, which plays a critical role in angiogenesis [71]. Cardiovascular complications are the primary cause of mortality and morbidity in T2DM. High certainty of evidence supported an antihypertensive effect of resveratrol supplementation among T2DM patients, with a reduction of 7.97 mmHg in systolic blood pressure and 3.55 mmHg in diastolic blood pressure compared with the placebo group. Moreover, a more pronounced reduction in blood pressure was observed in a subgroup with a shorter duration of resveratrol supplementation (< 12 weeks). In contrast, a longer duration of intervention did not affect blood pressure. In line with our study, previous research has pointed out that prolonged use of resveratrol may lead to some adverse effects on endothelial health, including elevation of CVD risk biomarkers such as soluble E-selectin 1 [72]. Additionally, in a study with patients undergoing peritoneal dialysis, the daily consumption of 450 mg of resveratrol over 12 weeks decreased vascular endothelial growth factor and angiopoietin-2 (angiogenesis markers) when the highest dose was ingested [73]. Thus, it was recommended that further studies should pay attention to the optimization of the dose and time frame of resveratrol use.

Moderate certainty of evidence from our UR indicated that resveratrol may have a mild protective effect on the kidneys in adults, including a decrease in BUN and serum creatinine levels and an elevation in glomerular filtration rate. Several previous studies have also shown similar results. For instance, a meta-analysis of 25 studies comprising data from 436 animals suggested that resveratrol treatment was significantly associated with lower levels of serum creatinine (SMD = -2.35, P < 0.0001) and BUN (SMD = -4.89, P < 0.0001), indicating that resveratrol has a beneficial effect on acute kidney injury (AKI) animal models [74]. Moreover, a recent animal experiment showed that resveratrol can aid in post-ischemic repair response and accelerate tubular recovery by reducing oxidative stress and mitochondrial damage [75]. In this context, Martínez-Rojas et al. also found that a 10-day treatment with resveratrol by gastric gavage (20 mg/kg/day) was sufficient to prevent the transition of AKI to chronic kidney disease (CKD), proposing resveratrol as a valuable candidate for further translational clinical studies to mitigate AKI-to-CKD transition [75]. Notably, these potential mechanisms are based on current scientific understanding and require further investigation to fully elucidate the underlying pathways.

Strengths and limitations

The main strength of this work is that we pool evidence from up-to-date published systematic reviews and meta-analyses on human health outcomes related to resveratrol and deliberately consider the convergence of results from different studies. Compared with the two previous URs considering the health effects of resveratrol [10, 11], our UR encompassed a wider spectrum of health indicators (including renal function, hormonal balance, and bone health, etc.), and examined resveratrol’s effects across diverse adult health statuses. Moreover, we carried out rigorous sensitivity analyses to determine the stability of the results, which was also lacking in the previous two URs [10, 11]. Of note, compared with the UR of Zeraattalab-Motlagh et al., we set a criterion of including only meta-analyses comprising three or more RCTs, thereby enhancing evidence precision [10]. In addition, given the scarcity of available observational studies identified during the screening process, the current UR was based on RCTs, which enabled the establishment of causation. Rigorous criteria, including AMSTAR and GRADE, were meticulously applied to evaluate the methodological quality of the included studies and assess the certainty of available evidence.

Despite the strengths, our UR also has some limitations that should be acknowledged. First, the number of included trials was low for some outcomes. For example, only three RCTs focused on the effect of resveratrol among patients with PCOS [76–78], which limited the generalizability of the conclusions. Secondly, the overlap of primary studies in the results can introduce redundancy and potentially skew findings [79]. However, in the current overview, the impact appears to be minimal due to eliminating overlap, ensuring that the results remain predominantly independent. Thirdly, some associations (54.8%) showed high or very high heterogeneity, as we used an I² > 50% and I² > 75% as a criterion for downgrading by one or two levels for inconsistency. Nevertheless, we conducted several sensitivity and subgroup analyses to clarify the source of the heterogeneity. Fourthly, harmful or less beneficial effects were observed with higher doses (≥ 500 mg/d) or longer duration (≥ 12 weeks) of resveratrol intervention in the subgroup analyses. However, we did not perform a quantitative analysis of the side effects of resveratrol supplementation, as there are not enough clinical studies that report this information. Previous reviews have reported some toxicity and adverse effects of resveratrol [80]. For instance, evidence suggests that the hormetic property of resveratrol may be due to its dose-associated biphasic effect on the cellular redox state, which was reported to be an antioxidant at low doses and a pro-oxidant at high doses [81, 82]. Fifth, we acknowledge that due to space limitations and the large number of non-significant associations identified, our results section primarily focused on findings that were both statistically significant and supported by moderate-to-high quality evidence. Although we have provided complete information on non-significant associations supported by moderate-to-high quality evidence (Supplementary Table 6), we recognize that this selective emphasis may introduce potential interpretation bias. Readers are therefore encouraged to interpret the main findings with caution and refer to the supplementary tables for a more comprehensive overview of the evidence. Lastly, the dependability of our study depended on the incorporated meta-analyses and the foundational studies they comprised. Although we traced the primary individual studies to confirm the accuracy of the data, there is still a possibility of errors beyond our ability to examine fully.

Conclusion

This UR has provided compelling evidence of the anti-obesity effects of resveratrol, including declining waist circumference and reducing the total cholesterol level among overweight adults, as well as the hypotensive effects of resveratrol among patients with T2DM, supported by high certainty of evidence. Moreover, several moderate-certainty evidence points to the positive effects of resveratrol supplementation on renal function, liver health, the hypoglycemic effect, and the therapeutic effect of inflammation and PCOS. Nevertheless, clinical trials with larger sample sizes and more extensive animal experiments are required to fully elucidate the mechanism and apply the protective properties of resveratrol in clinical practice. In addition, it was recommended that further studies with varying doses and durations should be carried out to ascertain the most efficient dose and duration for the protective effects of resveratrol.

Abbreviations

T2DM

Type 2 Diabetes Mellitus

CVD

Cardiovascular Disease

UR

Umbrella Review

MetS

Metabolic Syndrome

NAFLD

Non-alcoholic Fatty Liver Disease

RCT

Randomized Controlled Trial

BMI

Body Mass Index

LH

Luteinizing Hormone

SMD

Standardized Mean Difference

MD

Mean Difference

CI

Confidence Intervals

AMSTAR

A Measurement Tool to Assess Systematic Reviews

GRADE

The Grading of Recommendations, Assessment, Development, and Evaluation

PCOS

Polycystic Ovary Syndrome

TNF-α

Tumor Necrosis Factor-α

BUN

Blood Urea Nitrogen

AKI

Acute Kidney Injury

CKD

Chronic Kidney Disease

Authors’ contributions

Conceptualization: J-NS, RY, Q-JW, and T-TG. Data curation: J-NS, RY, LF, and Y-JL. Formal analysis: J-NS and LF. Writing – original draft & review and editing: J-NS, RY, LF, W-YX, M-LS, and LW. All authors read and approved the final manuscript. J-NS, LF, RY, and Y-JL contributed equally to this work.

Funding

This work was supported by the Natural Science Foundation of China (No. 82103914 to T-TG, and No. 82073647 and No. 82373674 to Q-JW) and Outstanding Scientific Fund of Shengjing Hospital (Q-JW).

Data availability

Data described in the manuscript, code book, and analytic code will be made available upon request.

Declarations

Ethics approval and consent to participate

The protocol for the UR was registered with the International Prospective Register of Systematic Reviews [PROSPERO (https://www.crd.york.ac.uk/PROSPERO/); registration number CRD 42024554480].

Not applicable.

Competing interests

The authors declare no competing interests.