The present descriptive qualitative systematic review followed the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) statement (20).

The search, conducted in July-August 2023, aimed to identify studies on the mechanisms of resveratrol in preventing PTB in animal models. The present systematic review followed the PICOS format: Population (P)-Pregnancy with induced PTB; Intervention (I)-Resveratrol; Comparison (C)-Negative control (untreated with resveratrol); Outcome (O)-Prevention of PTB (tocolytic effect); Study Design (S)-Preclinical studies. Specific search strategies were used for various databases, including MEDLINE via PubMed, ProQuest, EBSCOhost and EMBASE. The search queries and MESH keywords combined with Boolean operators were used, such as ‘Resveratrol’ (Mesh) AND [‘Premature Birth’ (Mesh) OR ‘Obstetric Labor, Premature’ (Mesh)].

The studies were selected based on the eligibility criteria using the PICOS strategy. The inclusion criteria were as follows: i) English-language articles; ii) primary experimental studies utilizing resveratrol (in vitro and in vivo); and iii) outcomes related to PTB. The exclusion criteria encompassed the following: i) Duplicated articles; ii) abstracts and reviews lacking specific data; iii) clinical studies, systematic reviews, case reports, retrospective studies, theses, letters, editorials, opinions, surveys, guidelines, conferences, abstracts and commentary articles. Initially, a search was performed for human population studies; however, as none were found, a focus was placed solely on reviewing experimental animal studies. Articles focusing solely on the benefits of resveratrol for preterm infants after birth were also excluded. In total, the authors (MH, MIDR and ABP) independently screened titles and abstracts, removing duplicates. Eligible articles had their full texts examined for confirmation. No restrictions on publication dates were applied to the systematic review. The search process is illustrated in Fig. 1.

To assess the quality of the studies, the ToxRTool (Toxicological Data Reliability Assessment Tool), developed by ECVAM (European Centre for the Validation of Alternative Methods) (21) was used. This tool provides criteria for evaluating the quality of pharmacology research using in vitro and in vivo experimental study designs. Previous systematic reviews that utilized the ToxRTool were also considered to determine the methodological reliability and inherent flaws in the studies (22-24). The ToxRTool consists of a 21-point rating checklist that assesses the methodological aspects of each study across five categories (25). These categories cover various aspects, including Category 1 (test substance identification), which includes details about (1.1) the substance, (1.2) purity, (1.3) source, and (1.4) nature and physico-chemical properties; Category 2 (test organism characterization), which includes (2.1) species (for in vivo studies) or test system (for in vitro studies), (2.2) sex (for in vivo studies) or the origin or source of the test system with the corresponding sex characteristic (for in vitro studies), (2.3) strain of test animals (for in vivo studies) or the specification of cell/tissue culture (for in vitro studies), (2.4) age or body weight of test organisms at the start of the study (for in vivo studies) or relevant clinical characteristics of human donors for cell/tissue cultivation (for in vitro studies), and (2.5) housing or feeding conditions in repeated dose studies (for in vivo studies) or cultivation process maintenance (for in vitro studies); Category 3 (study design description), which includes (3.1) administration method, (3.2) administered doses or concentration in the application media, (3.3) exposure frequency, duration, and explanation of observation time-points, (3.4) inclusion of negative and positive controls, (3.5) the number of animals (for in vivo studies) or the amount of cell/tissue culture per group (for in vitro studies), (3.6) administration scheme details for study evaluation, and (3.7) verification of achieved concentrations or substance stability in repeated dose studies; Category 4 (study endpoints), which includes (4.1) clear description and determination methods of endpoints, (4.2) transparent and comprehensive results description for all endpoints, and (4.3) transparent statistical methods for data analysis; and Category 5 (study design appropriateness and reliability), which includes (5.1) appropriate choice of study design for substance-specific data and (5.2) reliability of quantitative results. Studies with <13 points are deemed unreliable, those with 13-17 points are considered reliable with potential restrictions, and those scoring 18-21 points are deemed reliable without restrictions.

In addition, a risk of bias (RoB) assessment tool based on the Cochrane Handbook (26) was used with modifications that integrated the Office of Health Assessment and Translation, National Institute of Health (OHAT-NIH), and the National Toxicological Program tools, as used in previous studies (27-29). Three authors independently assessed all included studies for bias risk across nine parameters (randomization, allocation concealment, experimental conditions, blinding during the study, data completeness, exposure characterization, outcome assessment, selective reporting and data sufficiency). The responses were categorized as low risk (yes, provided information), unclear risk, or high risk (no, absent information) of bias following predefined protocol criteria. Any disagreements, conflicts, or discrepancies regarding the eligibility of studies were resolved through discussion and consensus to determine inclusion.

Data extraction was performed using Microsoft Office Excel 365. Three reviewers independently organized the data into variables based on the review's aims and topics. These variables included publication year, author(s), first author's country, study design, level of evidence (LoE) according to the Oxford Centre of Evidence-based Medicine (CEBM) criteria (30), preterm model sample, sample size, induced diseases and definitions. The outcomes data included the dose of resveratrol and administration details, treatment duration and frequency, additional drugs used, and associated study outcomes.

The relevant data of interest regarding the aforementioned variables were collected and organized into summary tables for study characteristics and outcomes using an electronic spreadsheet. The studies were further grouped based on similarities in reducing preterm rates and common underlying mechanisms.

The PRISMA flow diagram was followed for allocation concealment, blinding, article review, and data extraction. Initially, 3,597 results were retrieved from all databases, which were reduced to 2,791 after removing duplicates. Screening based on titles and abstracts resulted in 1,103 articles, and after applying inclusion and exclusion criteria, 20 full-text articles were assessed for eligibility. Ultimately, only four articles were eligible for analysis and review. Additionally, searches were conducted on Google Scholar and various Gray Literature websites (http://www.opengrey.eu/search/, https://v2.sherpa.ac.uk/opendoar/, https://www.worldcat.org/, and https://ukhsalibrary.koha-ptfs.co.uk/greylit/). However, no additional results were obtained beyond those included in the initial selection from the central journal databases.

A full-text assessment was conducted after removing all duplicate references using Zotero version 6.0.26 as the reference manager. The exclusion criteria included studies that focused on the effects of resveratrol on neonates as opposed to maternal conditions, such as retinopathy, lung injury, vascular dysfunction, embryopathy and oxidative stress-related disorders (n=7). Studies that discussed the benefits of resveratrol for cellular aging and senescence in cancer were also excluded (n=6). Additionally, a study that assessed the effects of resveratrol on non-pregnant myometrium was excluded (n=1), as well as narrative reviews (n=2).

As regards the reliability assessment (Table I), two studies [Bariani et al (8) and Novaković et al (31)] were deemed fully reliable and were included in the systematic review. However, two other studies [Deng et al (11) and Furuya et al (17)] were labeled as ‘reliable with restrictions’ due to methodological uncertainties and the absence of critical quality measurement categories. Specifically, the study by Deng et al (11) lacked information on the purity and origin of resveratrol, the age and other characteristics of the mice used, the environmental conditions for the mice, positive controls, a precise number of samples, and details on the stabilization and extraction/obtaining of resveratrol. Similarly, the study by Furuya et al (17) had limitations in sample characteristics and conditioning during the experiment, as well as a lack of a positive control and information on the stabilization of resveratrol. Detailed assessments for each category are provided in Table SI, Table SII and Table SIII.

In the risk of bias summary (Fig. 2), Bariani et al (8) had a low risk of bias in seven out of nine domains in each RoB category, with two aspects rated as high risk. Novaković et al (31) had a low risk of bias in five of nine domains in each RoB category, with one aspect considered unclear risk and the remaining three as high risk. Deng et al (11) and Furuya et al (17) each demonstrated a low risk of bias in three domains, an unclear risk in three other domains, and three domains with high risk. Despite all the limitations in the reliability and quality assessment, all studies were eligible to be assessed and reviewed in the present systematic review. According to the CEBM levels, they all fall under the fifth level of EBM (1 to 5).

The present systematic review included two experimental in vivo studies [Bariani et al (8) and Furuya et al (17)] and one study combining animal and cell experiments [Deng et al (11)]. All these studies investigated the effect of resveratrol on preventing PTB in animal models. Additionally, one in vitro study [Novaković et al 2015(31)] focused on the tocolytic and myometrium relaxant effects of resveratrol using human myometrium tissue to determine the preterm mechanisms. In this section, several studies that examined the preventive effects of resveratrol on PTB are presented. These studies include information about its characteristics (Table II), as well as its beneficial preventive effects on PTB (Table III).

The limited understanding of the complex etiology of PTB is a barrier to the development of effective strategies for its prevention. Recognized risk factors for PTB include multiple pregnancies, chorioamnionitis, maternal diseases, genetic factors, previous PTB and uterine abnormalities. However, not all cases of PTB have identifiable risk factors. The pathophysiology of PTB involves various processes, including myometrial contraction and extracellular matrix degradation. Among these processes, inflammation, which is initiated following implantation, plays a crucial role in PTB (37,38). Experimental animals, primarily mice or rats, play a critical role in studying the pathophysiology of PTB. The commonly used inflammation-inducing agent is LPS, derived from Escherichia coli endotoxins, administered intrauterine or intraperitoneally (39). In addition to LPS, various methods, including hormonal or immune agents, can induce PTB in mice (39).

Immunity, both innate and adaptive, plays a crucial role in the incidence of PTB. In mouse studies, macrophages induce PTB by producing pro-inflammatory cytokines such as matrix metallopeptidases, IL-1, IL-6 and IL-8. In the mouse model, PTB can be prevented by inhibiting the IL-1 receptor via the NF-κβ pathway (40). IL6-deficient mice exhibit resistance to LPS-induced PTB (41). Complement activation is also significant in preventing PTB, as women with spontaneous preterm labor have increased levels of complement C3a, C4a and C5a in infection-associated cases. Mice lacking the C5aR complement receptor demonstrate resistance to LPS-induced PTB in vitro (42).

In addition to the role of the innate immune system in PTB, the adaptive immune response is significant. PTB often occurs due to maternal intolerance to fetal antigens. Regulatory T-cells (Tregs) and effector T-cells (Teffs) undergo changes in frequency and phenotype. Tregs in women in preterm labor exhibit differential activation and reduced suppressive capacity (6). Toll-like receptors (TLRs) are another crucial component of the immune response. They activate signaling pathways that lead to cytokine and chemokine secretion by innate immune cells. TLR activation initiates inflammatory pathways that result in vaginal and preterm labor, immune cell recruitment, PG and MMP production, cervical ripening, and uterine contractions. TLR4 and TLR2 influence the timing of delivery during pregnancy (43,44). In addition to maternal TLR expression, fetal TLR expression may impact birth timing, with polymorphic alleles of fetal TLR4 and TLR2 associated with prematurity (45).

Current research explores the role of the microbiome role in modulating the risk of PTB. The composition of the vaginal microbiota influences the risk of PTB, with dysbiosis linked to increased levels of pro-inflammatory cytokines, such as CXCL10, in early pregnancy (46). However, administering pre-conception antibiotics to women with a history of PTB does not reduce the rates of PTB; instead, it is associated with a lower occurrence of LBW babies and early labor (47).

Among the various mechanisms associated with PTB, including innate immunity, adaptive immunity and dysbiosis, there is a notable interest in elucidating the mechanisms through which resveratrol intervenes in these processes. Resveratrol demonstrates antimicrobial properties against various pathogenic microorganisms, encompassing both Gram-positive and Gram-negative bacteria, as well as fungi. It can inhibit the growth of specific bacterial species at concentrations below 100 µg/ml, although higher concentrations are required to inhibit the growth of many bacterial organisms (48). The mechanism of action of resveratrol involves multiple pathways. It inhibits the electron transport chain and F0F1-ATPase, resulting in reduced cellular energy production. Resveratrol also interferes with DNA through the formation of a Cu(II)-peroxide complex and suppresses cell division by targeting the FtsZ gene. Moreover, it is effective in preventing biofilm formation and acts as both an antibiofilm and an anti-quorum sensing agent (49). The additional proposed mechanisms of action of resveratrol have been succinctly outlined in Fig. 3.

The beneficial effects of resveratrol in inflammation-induced preterm delivery involve multiple pathways. One of these pathways is the decrease in the activity of uterine NOS and the protein expression of uterine iNOS. The excessive production of NO leads to the formation of RONS, which are associated with tissue damage, oxidative and nitrative stress (50). To assess the activity of NOS and the protein expression of iNOS in the uterus, measurements were taken 5 h following the injection of LPS. The results indicated an increase in NOS and iNOS activity in mice treated with LPS alone. Bu contrast, mice treated with resveratrol after LPS exposure showed no significant increase in NOS activity (51). This reduction in uterine NOS and iNOS activity is beneficial as it prevents excessive production of NO and its associated adverse effects, such as damage to reproductive tissues and oxidative stress (51).

Reduction of levels of inflammatory cytokines (TNF-α, IL-1β and IL-6) in the cervix and peritoneum. Resveratrol is known to inhibit the activity of NF-κβ, a transcription factor involved in the expression of pro-inflammatory cytokines, in various immune response-related disease models (52). This factor operates not only in immune cells but also in the uterine myometrium, decidual stroma and amniotic cells (17). It promotes the production of cytokines, such as TNF-α, IL-1β, IL-6 and IL-8, which are crucial for inducing labor, particularly in preterm cases (53,54). Resveratrol appears to inhibit this NF-κβ/cytokine loop (17).

Furuya et al (17) focused on murine pro-inflammatory cytokines and observed significant increases in the levels of TNF-α, IL-1β and IL-6 in peritoneal washes and cervical tissues following exposure to LPS. In their in vivo model, resveratrol effectively suppressed the elevation of TNF-α and IL-1β levels induced by LPS, with a slight reduction in IL-6 levels (17). Although IL-6, produced by various stromal cells, was not as markedly suppressed, it may be due to its induction by other pro-inflammatory cytokines and direct stimulation by LPS (55). Since the preterm delivery model is initiated by the stimulation of TLRs in macrophages by LPS, inhibiting the expression of TNF-α and IL-1β in peritoneal washes and the uterine cervix directly suppresses this signaling pathway (17).

Resveratrol effectively suppresses the production of pro-inflammatory cytokines in the cervix, which is a key factor in the pathogenesis of PTB. PTB often involves local inflammation caused by microbial infection, resulting in elevated levels of cytokines, PGs, uterine contractions, and cervical ripening (56). In the study by Furuya et al (17), cervical tissues exposed to LPS in the control group exhibited significantly higher mRNA levels of TNF-α, IL-1β and IL-6 compared to the group treated with resveratrol(17). Resveratrol notably reduced the peak levels of TNF-α and IL-1β mRNA, particularly resulting in significantly lower TNF-α levels (0.54±0.09) in the resveratrol-treated group compared to the control group (2.45±0.93) (17). Resveratrol also demonstrates anti-inflammatory effects on placental tissues in pregnant women by downregulating the sirtuin-1 gene post-delivery (57). Its mechanism involves inhibiting the transcription activity of NF-κB, reducing the expression of the NF-κβ sub-unit RelA/p65, and stabilizing inhibitory I-κβ (57,58). Previous studies have demonstrated that resveratrol can suppress inflammatory responses in endometrial stromal cells in patients with endometriosis. The excessive production of pro-inflammatory cytokines can further stimulate the overproduction of PG, leading to increased uterine contractions and cervical maturation (59). Therefore, the ability of resveratrol to suppress pro-inflammatory cytokines in the cervix holds promising therapeutic implications for preventing PTB.

The pathogenesis of PTB involves both the cervix and the peritoneal area. In the study by Furuya et al (17), the injection of LPS into the cervix significantly increased the levels of pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6) in peritoneal washes, with IL-1β spiking to levels 480 times higher (17). This led to intrauterine inflammation and PTB in the experimental model. Peritoneal washes assessed the local immune responses to LPS, and the administration of resveratrol markedly decreased the levels of TNF-α and IL-1β. TNF-α levels were reduced to one-third, and IL-1β to about one-fourth, in mice treated with resveratrol compared to controls. However, there was no significant difference in IL-6 levels between the groups (17). Furuya et al (17) also demonstrated that resveratrol prevented the increased production of TNF-α and IL-1β by peritoneal macrophages in response to LPS. Comparing the local immune responses within 4 h of the injection, the results of peritoneal lavage revealed a significant decrease in TNF-α and IL-1β levels following the administration of resveratrol (17). TNF-α levels were approximately one-third in mice treated with resveratrol (16.7 vs. 53.9 pg/m), while IL-1β levels were approximately one-quarter (88.4 vs. 403 pg/ml, P<0.01). However, there was no significant difference in IL-6 levels between the resveratrol-treated mice and the controls (17).

Reducing inflammatory mediators from peritoneal macrophages. COX-2 enzymes are upregulated in response to pro-inflammatory cytokines, particularly in macrophages. This leads to elevated cytokine-mediated PGs, including PGE2 and PGF2α, which are significant in the mechanisms of PTB (60,61). TNF-α and IL-1β are central to the COX-2 pathway, and macrophages, responsible for their production, are suspected to play a role in the local inflammatory mechanism of PTB (8). In the study by Furuya et al (17), LPS-exposed peritoneal macrophages exhibited a significant increase in COX-2 mRNA levels for TNF-α and IL-1β. Resveratrol treatment effectively suppressed COX-2 production in these macrophages. This aligns with prior observations in peritoneal washes and cervical tissues by Furuya et al (17). Resveratrol reduced TNF-α, IL-1β, and COX-2 mRNA levels in macrophages induced by LPS in a concentration-dependent manner (17). Macrophages play a crucial role in COX-2 production and subsequent PGE2 and PGF2α production. A previous study demonstrated that omega-3 fatty acids, which also have antioxidant and anti-inflammatory properties, can suppress macrophage function and the elevation of PGE2 and PGF2α in gestational tissues induced by LPS (62). These findings collectively suggest that resveratrol primarily targets uterine-associated macrophages, offering the potential to modulate the local inflammatory response in PTB.

Reduction of COX-2 and PGs in the uterus. COX-2 produces PGE2 and PGF2α, potent inducers of uterine contractions and cervical ripening (19). PGs play a pivotal role in cervical maturation and uterine contractions, increasing amniotic fluid during labor (63,64), an essential process in childbirth (65). The upregulation of COX-2 is associated with PTB pathology in both mouse models and humans. The study by Bariani et al (8) revealed that LPS-induced uterine inflammation increases COX-2 expression, initiating preterm labor. This upregulation occurs in tissues such as the amnion, choriodecidua, and myometrium, linking it to PTB pathology (66).

The ability of resveratrol to modulate COX-2 expression, a key player in PG production, led to the assessment of uterine PG levels, PGE2, PGF2α, 6-keto-PGF1α and PGI2(8). Resveratrol has been shown to effectively suppress the LPS-induced elevation of COX-2 in gestational tissues (8). Its administration results in the downregulation of both COX-2 mRNA and protein expression in the uterus, preventing COX-2 gene and protein overexpression induced by LPS (8,67). Notably, LPS alone or in combination with resveratrol does not alter the levels of 15-hydroxyprostaglandin dehydrogenase (15-Pgdh) mRNA and COX-1 protein levels (8), which decrease PG production, counteracting the effects of COX-2(68). In the study by Bariani et al (8), resveratrol administration prevented the increase in uterine PGE2 levels and partially affected PGF2α levels, with no significant impact on the 6-keto-PGF1α concentration (8). The reduction in pro-inflammatory cytokines by resveratrol may explain the lower PG expression, as TNF-α and IL-1β induce PGE2 and PGF2α in choriodecidual tissues (69). In summary, by decreasing PG synthesis, resveratrol exhibits potential in helping to prevent preterm labor.

Reduction of endocannabinoid levels. The endocannabinoid system (ECS) consists of cannabinoid receptors CB1 and CB2, endogenous lipid ligands called endocannabinoids, and regulating enzymes that synthesize and break down these lipids. All endocannabinoids belong to the eicosanoid class (70). Endocannabinoids are unsaturated fatty acid derivatives that interact with this system, eliciting specific responses (71). Anandamide (AEA) and 2-arachidonoylglycerol (2-AG) are key endocannabinoids that are released enzymatically from cell membrane phospholipid precursors in response to various stimuli, including neurotransmitters, depolarizing agents, and hormones (71). AEA is present in the uterus, and its levels in the blood increase during normal labor, suggesting the involvement of the ECS in reproductive and preterm labor processes (72-74). The study by Bariani et al (8) demonstrated that LPS and/or resveratrol affected uterine endocannabinoid levels. LPS did not affect the 2-AG levels, but resveratrol alone reduced its levels. AEA increased with LPS, but resveratrol prevented this increase.

The effects of AEA on the uterus are conflicting, with some studies demonstrating the relaxation of the pregnant myometrium (75), while others have reported that the activation of CB1 receptors leads to the increased production of PGE2 in the amnion and choriodecidua (76). CB1 receptor activation is associated with the increased production of PGF2α in inflammation-induced preterm labor (77). Prolonged endocannabinoid signaling is linked to adverse pregnancy outcomes (78). Bariani et al (8) found that LPS altered AEA and related endocannabinoid-like lipid levels in the uterus, consistent with previous research (75,76). LPS-induced cytokines can affect endocannabinoid production by modulating the expression and activity of fatty acid amide hydrolase (FAAH) (79,80). In the study by Bariani et al (8) treatment with resveratrol restored AEA and other lipid levels to the control levels, although the exact mechanisms involved remain unclear. The increased AEA production induced by LPS was associated with a higher incidence of preterm labor, while the tocolytic effects of resveratrol correlated with reduced uterine AEA content (8).

Bariani et al (8) discovered that LPS increased the AEA levels in the uterus, resulting in a greater likelihood of preterm labor. However, resveratrol prevented this increase, indicating its tocolytic effect in reducing uterine endocannabinoids during inflammation (8). In addition to endocannabinoids, endocannabinoid-like lipids from the N-acyl ethanolamine, N-arachidonoyl glycine (NAGlys), N-arachidonoyl serine (NASers) and 2-acyl-sn-glycerols families can undergo changes in composition during LPS-induced pregnancy inflammation (8). The specific molecular mechanism through which resveratrol modifies uterine endocannabinoids and lipid composition remains unclear. However, it is known that NAGlys and NASers, which are part of these lipids, do not activate CB1 or CB2 receptors (81). NASers have been observed to regulate calcium-activated potassium channels (82) and N-type calcium channels (83). Additionally, NAGlys can be metabolized by COX-2 and FAAH, potentially functioning as endogenous enzyme inhibitors (84).

There was a 1.5-fold increase in the average plasma levels of AEA from 1.20±0.57 nM in non-labor conditions to 1.82±0.87 nM during labor (P<0.0001) (72). A plasma AEA concentration exceeding 1.095 nM was found to be a better predictor of PTB before 37 weeks of gestation compared to currently used predictors, such as cervical length, oncofetal fibronectin and cervical insulin-like growth factor binding protein 1. It demonstrated 87.14% specificity, 25.93% sensitivity, a 70.24% negative predictive value (NPV), and a 61.2% positive predictive value (PPV), surpassing existing predictive tests (85). The average plasma AEA level significantly increased in women with a positive fetal fibronectin test compared to those with a negative result, providing additional evidence of the role of AEA in labor and its potential use as a predictor for PTB (85). AEA levels are regulated through cellular uptake by the AEA transporter and enzymatic degradation by FAAH on the cell membrane (86). LPS-induced cytokines influence endocannabinoid production by affecting FAAH protein expression and enzymatic activity (87). Anti-inflammatory cytokines (IL-4 and IL-10) enhance FAAH activity, while pro-inflammatory cytokines (IL-12 and IFN-γ) reduce FAAH activity and protein expression in peripheral mononuclear cells (lymphocytes) (8,88). During pregnancy, FAAH levels in peripheral mononuclear cells show an inverse correlation with blood AEA levels (89). Low AEA levels with high FAAH are crucial for the successful development of pregnancy (90,91). The reduced activity and expression of anandamide hydrolase in peripheral lymphocytes can serve as an early indication of spontaneous abortion before eight weeks of gestation, offering potential diagnostic utility for pregnancy monitoring and fertility assessment (91).

Modulating 5' AMP-activated protein kinase (AMPK) and mammalian target of rapamycin complex (mTORC) pathways. Inflammation and oxidative stress are widely recognized risk factors for PTB. Studies conducted on mouse models, specifically those with spontaneous and inflammation-induced PTB with p53 deletion (p53^d/d females), have indicated the involvement of mTOR activation in decidual senescence and the excessive production of PGs derived from COX2, ultimately leading to PTB (11). In female mice with the p53^d/d phenotype, a noticeable decrease in the growth of uterine decidual cells due to accelerated senescence has been observed (92,93). This condition is accompanied by elevated levels of pAkt, p21, IL-8 and various CXCL cytokines, collectively contributing to the senescence-associated secretory phenotype, which further promotes decidual senescence and PTB (93). Additionally, this p53 deletion has been associated with PTB, which can also be influenced by factors, such as increased ROS production due to conditions such as pre-eclampsia, as well as environmental factors such as smoking and pollution (93).

Deng et al (11) demonstrated that resveratrol, an antioxidant and anti-aging agent, activated AMPK signaling and inhibited mTORC1 signaling in decidual cells without causing any adverse effects on the mother or offspring. Consequently, resveratrol offers protection against both spontaneous and inflammation-induced PTB in p53^d/d females. Similar beneficial effects have also been observed with the use of AMPK activators such as metformin, which enhances decidual health and reduces PTB rates in mouse models (11). Human fetal membranes contain both AMPK and its phosphorylated form (p-AMPK). Membranes obtained from spontaneous term labor and pre-labor rupture display significantly lower levels of p-AMPK compared to intact membranes (11). Pre-treatment of human fetal membranes with AMPK activators reduces the production of inflammatory cytokines in response to LPS, indicating the anti-inflammatory effects of p-AMPK on fetal membranes (94). Resveratrol decreases PGE2 levels by suppressing Ptgs2, the COX2-encoding gene, in decidual cells. It also inhibits hyperactive mTORC1 through both AMPK-dependent and/or AMPK-independent pathways, underscoring its role in the regulation of AMPK and mTORC1 signaling in an inverse manner (95).

Other potential benefits of resveratrol: Smooth muscle relaxation. Resveratrol has the ability to relax smooth muscle, which may have an impact on PTB by targeting potassium (K⁺) channels (31,96-98). The proper functioning of K⁺ channels is crucial for maintaining a relaxed uterine state during pregnancy by aiding in membrane repolarization (99). The dysregulation of K⁺ channels can lead to abnormal uterine activity, contributing to conditions such as preterm labor (31). Poorly functioning or less active K⁺ channels can decrease the repolarizing current in myometrial smooth muscle cells, resulting in untimely uterine contractions and preterm delivery. Conversely, excessive expression of K⁺ channels in the later stages of pregnancy may hinder synchronized uterine muscle activity needed for full-term labor (98). Various types of K⁺ channels, including big Ca^2+-sensitive K⁺ (BKCa) channels, ATP-sensitive K⁺ (KATP) channels, Kv channels and SK channels, induce relaxation in both non-pregnant and pregnant myometrium, demonstrating the complex regulation of uterine tone (31,96-98). BKCa channels, which hold a particular influence on uterine smooth muscle, play a significant role in inducing smooth muscle relaxation and repolarizing K⁺ current (99). Their activity varies throughout pregnancy, and their modulation is regulated by 17b-estradiol (31). BKCa channels have a more pronounced relaxant effect during mid-gestation (100,101), indicating their importance in maintaining uterine quiescence (99).

The regulation of myometrial tone during pregnancy involves the modulation of ion channels. β-receptor agonists, used to delay preterm labor, interact with BKCa channels via ß2- and ß3-receptors, resulting in uterine relaxation (99). Human studies suggest increased levels of BKCa channel in late pregnant non-laboring myometrium but decreased expression in preterm and term laboring myometrium (99). KATP channels, composed of Kir6 channels and sulfonylurea receptors, play a role in myometrial quiescence; however, their specific function in the myometrium is not yet clear. The downregulation of KATP channels, particularly Kir6.1 and SUR2B subunits, may increase uterine excitability and labor contractions in term pregnancy (101). The regulation of KATP channels varies with the stage of gestation and the incidence of labor contractions in the human myometrium, necessitating further research for a comprehensive understanding (101). Aberrant K channel function or expression is associated with conditions such as PTB and preeclampsia (96,102). Resveratrol can modulate K channel activity, potentially offering new treatments for these conditions by reducing intracellular calcium levels and inhibiting various contractions (31). Resveratrol effectively inhibits uterine smooth muscle contractions, which can lead to infertility, endometriosis, miscarriage, or PTB (31,96,97,99). It targets KATP, BKCa and Kv channels to achieve relaxation effects on the uterus (97). This relaxation potential extends to pregnancy-related disorders, such as PTB, by relaxing the myometrial smooth muscle and fetoplacental blood vessels, inhibiting contractions induced by oxytocin, PGs and acetylcholine (96).

Resveratrol shows promise as an herbal preventive agent for PTB. In the study by Bariani et al (8), the administration of resveratrol resulted in a 64% reduction in PTB rates and a 28% decrease in mortality rates. These effects may be attributed to resveratrol's impact on uterine NOS activity, iNOS expression, COX-2, PGE2, and AEA profiling. Additionally, this treatment led to a decrease in antepartum hemorrhage, fetal toxicity, and offspring mortality. Importantly, mice treated with resveratrol did not experience vaginal bleeding, and their uterine morphology improved (8). Deng et al (11) demonstrated that resveratrol has the potential to prevent PTB through its modulation of AMPK and mTORC1 signaling in decidual cells. The administration of resveratrol offers several advantages, including no adverse effects on pregnancy outcomes, effective targeting of decidual health, and compensation for ovarian luteolysis (11). The study conducted by Furuya et al (17) also supports the tocolytic effects of resveratrol by suppressing peritoneal macrophage-induced inflammation and preventing spontaneous PTB in a genetic model of accelerated decidual senescence.

Furthermore, Furuya et al (17) discovered that resveratrol had a significant impact on reducing the rate of PTB in mice. At a dosage of 20 mg/kg, the PTB rate was 48.6±19.4% compared to 97.8±1.9% in the control group (P=0.03). Morever, at a dosage of 40 mg/kg, the PTB rate was 57.1±13.8% compared to 83.7±13.2% in the control group (P=0.15). By orally administering resveratrol, the occurrence of LPS-induced prematurity in mice was reduced through the suppression of inflammatory mediators, PGs, peritoneal macrophages and endocannabinoid compounds in the maternal reproductive organs (17). Novaković et al (31) demonstrated that resveratrol effectively prevents PTB by inhibiting uterine contractions, including those stimulated by oxytocin. Resveratrol achieves this by regulating various myometrial K⁺ channels, such as KATP, Kv and BKCa channels, by increasing intracellular Ca²⁺ levels. Moreover, it induces partial relaxation of the myometrium by activating Kir6.2/SUR1 channels. Nonetheless, there are certain limitations to consider, such as the failure of a single resveratrol treatment to prevent PTB in the study conducted by Deng et al (11) and its lack of effectiveness against tonic contractions induced by oxytocin (31).

Oral resveratrol is readily absorbed but undergoes rapid metabolism (102). Azachi et al (103) conducted a study in which red grape cell (RGC) resveratrol exhibited two plasma concentration peaks: One at 1 h following administration and another at 5 h following administration. The initial peak is likely a result of the extensive glucuronidation and sulfation of resveratrol in the enterocytes of the small intestine. By contrast, the second peak occurs due to the flow of bile-containing metabolites from the liver to the intestines, a normal occurrence after food consumption that may prolong the pharmacological effect of certain substances and their metabolites (103,104).

The median time to peak drug concentration (T_max) for total resveratrol at a dose of 150 mg RGC is 4 h (range, 0.67-6.00), and at 50 mg it is 1 h (range, 0.33-8.00). Moreover, the median T_max for free resveratrol at a dose of 150 mg RGC is 1 h (range, 0.33-4.00) (103). Resveratrol exhibits a high oral absorption but rapid metabolism, resulting in poor overall bioavailability. Even with a high dose of 5 g, it only reaches a peak plasma-free concentration of 538 ng/ml. This limited bioavailability may be attributed to low solubility, similar to other polyphenol aglycones, which struggle to form hydrogen bonds due to hydrophobic interactions with hydroxyl or aromatic groups (103). One approach to enhance solubility is through glycosylation of the resveratrol parent compound (103).

Resveratrol interacts with cytochrome P450 enzymes, inhibiting CYP3A4, CYP2D6 and CYP2C9, while inducing CYP1A2, potentially affecting drugs metabolized by these enzymes (105). Urinary excretion studies demonstrate high absorption rates (at least 70%) after oral resveratrol intake, with 53.4 to 84.9% of total radioactivity recovered in urine. However, rapid biotransformation leads to minimal unmodified resveratrol in circulation and relatively high levels (maximum, 2 µM) of resveratrol metabolites after a 25 mg oral dose, resulting in near-zero resveratrol bio-availability (104). The intravenous injection of a 0.2 mg dose shows no second peak and a rapid decline in plasma total radioactivity over the first hour, indicating extensive distribution. Plasma half-life durations are similar after both routes (9.2 h orally and 11.4 h intravenously), with dose-dependent resveratrol plasma levels reaching up to 530 ng/ml with higher doses (up to 5 g) (104). The administration of resveratrol during pregnancy appears to be relatively safe for both the mother and fetus, with no reported toxicity or adverse pregnancy outcomes (11,102).

Resveratrol is primarily sourced from grapes, grape skins, peanuts, red wine, cranberries and Japanese knotweed (Polygonum cuspidatum) (102,105). The quantity of resveratrol varies among these sources. For example, RGCs contain between 726 to 916 mg/kg (103), while agricultural red grapes range from 0 to 42.5 mg/kg (103). Other examples include jamun fruit with 11.19 to 34.87 mg/g of dry weight, mulberries with 35.8 to 50.61 mg/g, jackfruit with 0.07 to 3.56 mg/g (106), apples with 47 mg/g, broccoli with 18 mg/g, onions with 18 mg/g and black tea with 11.3 mg/g (107). Typical concentrations of resveratrol in various food products to meet the recommended daily allowance (RDA) of 1 g are as follows: Red grapes (92-1,604 mg/kg fresh weight), white grapes (59-1,759 mg/kg fresh weight), peanuts without seed coats (0.03-0.14 mg/g), red wines (0.361-1.972 mg/l), white wines (0-1.089 mg/l), rosé wines (0.29 mg/l), beers (1.34-77.0 mg/l), tomato skins (19 mg/g dry weight), dark chocolate (350 mg/kg), milk chocolate (100 mg/kg), Itadori tea (68 mg/100 ml) and apples (400 mg/kg fresh weight) (108).

Due to its limited solubility in water, resveratrol is primarily absorbed through passive diffusion in the colon and undergoes significant pre-systemic metabolism in the liver. This liver metabolism markedly reduces the levels of free resveratrol before it enters systemic circulation. Resveratrol is conjugated to more soluble glucuronides (e.g., resveratrol-3-O-glucuronide, resveratrol-4-O-glucuronide) and sulfates (e.g., resveratrol-trisulfate) in phase II metabolism or binds to albumin and lipoproteins. As a result, the therapeutic effectiveness of oral resveratrol is relatively low. Efforts to improve the bioavailability of resveratrol or delay its phase II metabolism through the use of resveratrol prodrugs are crucial for its biomedical applications (108,109).

The oral supplementation of 80 mg resveratrol in overweight pregnant women has demonstrated a reduction in the incidence of gestational diabetes mellitus, improvement in lipid profiles, and a decrease in glucose levels within a 60-day period. Additionally, resveratrol administration at a dose of 50 mg for up to 5 doses has been found to lower blood pressure in patients with preeclampsia (102). However, it should be noted that there is limited clinical research available on the use of resveratrol during pregnancy, with more extensive studies available for patients with diabetes, cardiovascular issues and metabolic disorders (105). Challenges in the clinical application of resveratrol include the wide range of dosages used (ranging from 5 mg to 5 g daily), the diverse dose-effect associations observed in the SIRT1 pathway, concerns regarding renal toxicity associated with micronized oral formulations such as SRT501 (commercially marketed as a supplement, rather than a definitive treatment drug) and gastrointestinal side-effects at high doses. Furthermore, it has been observed that resveratrol can activate estrogen-regulated genes that are associated with estrogen-dependent neoplasms (105).

This presented systematic review highlights a unique academic perspective on the prevention of PTB using resveratrol. It offers original contributions to the field of preclinical investigations of natural compound medicine for PTB. Notably, to the best of our knowledge, this systematic review is the first to specifically examine the impact of resveratrol in preventing PTB and highlights its promising benefits. The present study goes beyond a mere summary of the literature by providing a comprehensive framework that explains the preventive mechanisms of resveratrol. It also includes a visualized image that consolidates the scattered literature. By adopting this approach, the present study contributes significantly to the development of strategies to combat PTB. Furthermore, the present study stands out for its meticulous quality assessment, employing stringent predetermined criteria to critically evaluate the included studies.

In the present systematic review, the authors were not able to conduct a meta-analysis due to two main reasons. First, there was a limited number of articles included, which makes it impractical. Second, the data and results presented in these studies were presented in a heterogeneous manner, rendering it impossible to pool the data. The main limitation of this systematic review was the diverse nature of how data and results are presented across the included studies, which prevents a meta-analysis from being practically feasible. Additionally, while the significant prevalence of favorable outcomes in the present systematic review is encouraging, it is important to consider the possibility of publication bias, where studies with unfavorable results may remain unpublished. Furthermore, the strict inclusion criteria resulted in the exclusion of several studies that did not directly involve resveratrol in PTB. Nonetheless, the insights from these excluded studies are still integrated into the discussion section, adding to the broader conversation on the subject.

Despite the potential of resveratrol as a tocolytic agent for preventing PTB, particularly in cases involving intrauterine inflammation, there is a scarcity of human studies evaluating its effectiveness in this regard, and no current human clinical trials have been conducted, at least to the best of our knowledge. While mouse models provide valuable information, their limitations stem from significant anatomical and physiological differences compared to humans, and findings from in vitro and in vivo studies may not directly translate to human outcomes due to inherent species disparities. However, this juncture presents an opportune moment for researchers to validate the potential benefits of resveratrol through dedicated human studies. The current insights from various models offer valuable information about labor and delivery mechanisms, underscoring the need for well-designed prospective studies to assess the efficacy of resveratrol in human populations.

In conclusion, the findings presented herein indicate that resveratrol may have the ability to prevent preterm labor induced by LPS by inhibiting pro-inflammatory mediators, reducing PG levels, improving endocannabinoid profiles, relaxing smooth muscle and preventing excessive uterine contractions through ion channels. Considering its ability to cross the placenta, future studies are required to investigate its potential protective effects on embryos. Moreover, a better understanding of the role of resveratrol in human or similar uterine physiology necessitates further investigation. Urgent clinical trials are necessary to determine the optimal dosage of resveratrol for PTB therapy using standardized formulations. Given the low oral bioavailability of resveratrol, improving its pharmacological properties is pivotal. Structural optimization and the use of resveratrol-encapsulated nanoparticles can enhance efficacy, reduce dosages, minimize side-effects and target specific organs. Additionally, exploring the potential of locally derived resveratrol from plants and conducting oligomer research shows promise. Long-term studies are required to substantiate the scientific potential of resveratrol.