The effects of pro-, pre-, and synbiotics supplementation on polycystic ovary syndrome: an umbrella review of meta-analyses of randomized controlled trials

Background Synbiotics, refer to a combination of probiotics and prebiotics in a form of synergism that beneficially affect the host’s health by alternating the composition and/or function of the gut microbiota. Numerous meta-analyses of randomized clinical trials have proven that pro, pre-, and synbiotics supplementation has health outcomes in women with polycystic ovary syndrome (PCOS). However, the strength and quality of this evidence in aggregate have not yet been synthesized in great detail. Methods PubMed, Scopus, Web of Sciences, and Google Scholar were searched up to March 2023. We pooled the mean difference and its 95% confidence interval (CI) by applying a random-effects model. Results Overall, nine meta-analyses including a total of 12 trials were identified. The results of the present study indicated that probiotic supplementation significantly reduced the homeostatic model assessment for insulin resistance (HOMA-IR; WMD: −0.29, 95% CI: −0.57 to −0.02, p = 0.03, n = 4; moderate certainty) and fasting glucose concentration (FGC; WMD: −7.5 mg/dL, 95% CI: −13.60 to −0.51, p = 0.03; n = 4; low certainty). Moreover, synbiotic supplementation had beneficial effects on glycemic control, lipid profile, and hormonal parameters, but the certainty of the evidence was rated as low to very low. However, supplementation with pro−/synbiotics did not affect inflammation and oxidative stress in women with PCOS. Furthermore, waist/hip circumference, fasting glucose concentration, lipid profile, dehydroepiandrosterone sulfate, high-sensitivity C-reactive protein, and hirsutism score were significantly reduced after prebiotics supplementation with low certainty of evidence. Conclusion Although pro-, pre-, and synbiotics supplementation had beneficial effects on some PCOS-related outcomes, the certainty of the evidence was rated as low to very low. Therefore, further well-designed RCTs might help to confirm our findings in women with PCOS.


Introduction
Polycystic ovary syndrome (PCOS) is a common endocrinopathy that affects women of reproductive age, particularly in the early to late reproductive stages (15-35 years) (1,2).As defined in 2003 by the Rotterdam Consensus Declaration, the onset of two out of these following features is a sign of PCOS: oligo or anovulation, hyperandrogenism, and polycystic ovaries (3,4).Depending on diagnostic criteria it is estimated that between 5 and 21% of women worldwide are affected by PCOS (5).Major complications of PCOS include insulin resistance (IR), glucose intolerance, type 2 diabetes mellitus, dyslipidemia, cardiovascular disease (6), hirsutism (7), acne, alopecia (8), and high C-reactive protein (9).The financial burden of PCOS, including the costs of initial diagnosis and reproductive endocrine complications, was estimated at $ 3.7 million per year in the United States and taking into account the cost of pregnancy-related and long-term complications, it has risen to $8 million per year (10).
Multiple pathophysiological mechanisms are assumed due to the heterogeneity of the PCOS characteristics.Hyperinsulinemia and insulin resistance, exaggerated LH pulse frequency and amplitude, and enhanced ovarian or adrenal androgen production, are the main presumed causes of PCOS (11,12).
Recent studies regarding probiotics, "live microorganisms which when administered in adequate amounts confer a health benefit on the host, " demonstrated that the administration of probiotics can decrease intestinal permeability, modify the immune system of the gastrointestinal tract and prevent the growth of pathogenic bacteria (13)(14)(15).The term prebiotic is used as "a substrate that is selectively utilized by host microorganisms conferring a health benefit" (16).Short-chain fatty acids (SCFAs) from the metabolism of prebiotics, decrease inflammatory markers and subsequently reduce insulin resistance (17).The presence of a combination of living microorganisms and substrate(s) that host microorganisms use to their advantage and which benefits the host's health is called synbiotics (18).Synbiotics administration was associated with significant improvement in fasting plasma glucose (FPG), homeostatic model assessment for insulin resistance (HOMA-IR) and body mass index (BMI) (19).
A substantial number of systematic reviews and meta-analyses (SRMAs) of randomized controlled trials on the effects of pro-, pre-, and synbiotics supplementation on PCOS-related outcomes (6,(20)(21)(22) have been conducted in recent years.Regardless of the high number of SRMAs, there is still some uncertainty about the efficacy of each prebiotic, probiotics, and synbiotics supplement separately.There is also currently no available data to support the certainty of the evidence for each estimate and the amount of impact detected based on the minimal clinically important differences (MCID).Also, the strength and quality of this evidence in aggregate have not yet been synthesized in great detail.Therefore, this umbrella review aims to examine systematic reviews to determine the effectiveness of pro-, pre-, and synbiotics on hormonal parameters, glycemic control markers, blood lipids, anthropometric indices, and inflammatory and oxidative stress biomarkers in women with PCOS and update the evidence.

Methods
The current umbrella review was designed based on the protocols of the Cochrane Handbook for Systematic Reviews of Interventions on overviews of systematic reviews (23).The protocol of this umbrella review was registered in the International Prospective Register of Systematic Reviews (PROSPERO) database (https://www.crd.york.ac.uk/PROSPERO, CRD42021281029).

Search strategy
The systematic search was conducted in major databases including PubMed, Web of Science, Scopus, and Google Scholar until 22 March 2023, with no restrictions on publication time or language.Detailed information relating to the search strategy of databases as well as the medical subject headings (MeSH) and text words in our search strategy to identify relevant studies are provided in Supplementary Table 1.We also added other literature that was found by manually reviewing related published SRMAs of RCTs evaluating the effects of pro-, pre-, and synbiotics supplementation in women with PCOS.Moreover, the references list of any related meta-analyses was manually reviewed to collect further eligible studies.

Eligibility and study selection
Relevant studies were selected based on the PICOS (population/ intervention/comparison/outcome) framework: P (women with polycystic ovary syndrome), I (pro-, pre-and synbiotics supplementation), C (placebo), O (PCOS-related outcomes), and study design (SRMAs of RCTs).Two authors (ST and NP) independently selected meta-analyses in this umbrella review if they met the following criteria: (1) SRMAs of RCTs that were conducted in the people of any age with a diagnosis of polycystic ovary syndrome; (2) received at least one oral probiotic, prebiotic, or synbiotics supplementation compared to a control group; (3) reported weighted We excluded studies with insufficient data and other study designs.We also excluded primary trials in the meta-analysis if they: (1) were trials without a control group; (2) used pro-, pre-, and synbiotics supplementation in combination with other nutrients.If more than one published metaanalysis for a given outcome was available, we selected only the publication with the higher number of primary trials (24).Also, we have manually reviewed the reference lists of other meta-analyses to identify additional relevant trials.

Data extraction
NP extracted the following data from eligible meta-analyses using a pre-designed abstraction form: first author's name, country, publication year, number of primary studies, and participant number.Furthermore, for each primary RCTs from included meta-analyses, we also extracted the following required data: First author, country, publication year, effect size, participant number, duration of intervention, and the dose of supplementation.

Assessment of methodological quality
A Measurement Tool to Assess Systematic Reviews (AMSTAR-2) scale (25) was used to evaluate the methodological quality of included meta-analysis by two independent researchers (ST and SZM).Disagreements were resolved by consensus with the third researcher (SSH).We also carried out the quality of primary trials including each eligible meta-analysis using the Cochrane risk-of-bias tool for randomized trials (RoB) (26).According to this systematic bias assessment, the overall quality of primary studies was scored as good, fair, or weak (Supplementary Table 2).
The AMSTAR 2 tool (25) was applied to assess the quality of conduct of the included meta-analyses of randomized controlled trials.Instrument (AMSTAR 2) retains 10 of the original domains, and has 16 items in total.

Data synthesis and statistical analysis
For each health outcome, the largest meta-analysis with a maximum number of RCTs was selected, as well as primary trials that were ignored in the biggest meta-analyses were also added (Table 1).Then, we recalculated the MD and its 95% CI by applying a random-effects model in each meta-analysis that was included in our umbrella review (27).To evaluate the possibility of publication bias, we used Egger's test method (28).Heterogeneity across studies was estimated by Cochran Q and I 2 statistics, in which I 2 values greater than 50% or p < 0.05 were considered as significant (29).Statistical analyses were conducted using STATA version 14 software (Stata Corp, College Station, Texas, United States).

Grading of the evidence
The certainty of the evidence was rated according to the Grading of Recommendations Assessment, Development and Evaluations (GRADE) (30).The GRADE consists of five domains: risk of bias in the individual studies, inconsistency, indirectness, imprecision, and publication bias.As a result, high, medium, low, or very low-GRADE ratings were considered for the certainty of evidence.The MCID for the estimations was determined using previous data in the literature, and in the absence of sufficient evidence, we used half of the baseline SDs for that outcome (31).Supplementary Table 3 demonstrates the MCID values utilized in the current umbrella review.

Literature search
We identified a total of 91 meta-analyses studies through initial electronic searches.After removing 17 duplicated studies, 62 publications were assessed based on reviewing titles and abstracts.Of those, 12 records remained for full-text revision.Among them, three articles were excluded due to the full text being unavailable (32) and performed on other patients (33,34).Overall, nine meta-analyses were finally included in this umbrella review.The flow diagram of the study selection process is illustrated in Figure 1.Through the screening primary studies of included meta-analyses, five RCTs were excluded for either of the following reasons: full text being unavailable (n = 2) (35,36) and using probiotics in combination with other interventions (n = 3) (37)(38)(39).Detailed reasons for the exclusion of primary trials by full-text assessing are provided in Supplementary Table 4. Overall, nine meta-analyses (6,(20)(21)(22)(40)(41)(42)(43)(44) reporting 12 RCTs (45-55) met the eligibility criteria for the final analysis in this umbrella review.

Methodological quality
According to AMSTAR 2 scores, two meta-analyses were classified as high-quality studies (6,42), four meta-analyses were performed with a low-quality method (21,22,40,44), and the other three metaanalyses were performed with a critically low-quality method (20,41,43).Detailed AMSTAR scores for each meta-analysis are presented in Supplementary Table 5.

Publication bias
We found statistically significant publication bias regarding the levels of HDL-C (Egger's = 0.01) following intake of probiotic supplementation, and the levels of FGC after supplementation with synbiotics (Egger's = 0.04).Therefore, we did the trim-and-fill method to detect sources of bias and found results similar to the original.No evidence of publication bias based on Egger's tests was observed in other outcomes (Tables 3-5).

Discussion
The present work was performed on meta-analyses of RCTs to comprehensively assess the effects of pro-, pre-, and synbiotics supplementation on PCOS-related outcomes.We evaluated the evidence using the well-known GRADE tool and to provide better comparisons between outcomes, the available data were reanalyzed using random effects analysis.Our findings are important because there is limited evidence-based support for use of pro-, pre-, and synbiotics supplements in the management of PCOS-related outcomes.
The results of the present study showed probiotic supplementation significantly reduced HOMA-IR, FGC, and VLDL.In addition, synbiotics supplementation was found to have beneficial effects in the reduction of WC, fasting insulin, HOMA-IR, TG, VLDL, TC, LDL-c,  TT, and hirsutism score.Moreover, we found prebiotic supplementation significantly reduced WC, HC, FGC, TG, TC, LDL-c, dehydroepiandrosterone sulfate, hs-CRP, and hirsutism score.In contrast, our study showed that probiotic supplementation significantly increased WC, SHBG, TAC, and MDA parameters.It was also found synbiotics supplementation significantly increased SHBG and QUICKI.These findings should, however, be interpreted with some caution due to the following reasons: Firstly, almost all of the significant findings in the analyses received low and very low-quality evidence based on the GRADE tool.Only moderate quality of evidence was found for the effects of probiotics on the HOMA-IR index.None of the included meta-analyses considered this critical point and their findings were judged based on statistical differences.The included meta-analyses in this umbrella review were also evaluated for methodological accuracy using the AMSTAR tool.
According to this method, three meta-analyses showed critically low quality, three showed low quality, and two showed high quality.The meta-analyses were rated as low and critically low-quality methods because did not register the protocol of the meta-analysis, had no comprehensive search strategies, did not report the reasons for excluded studies, and did not discuss the possible risk of bias in primary studies.Secondly, most of the analyses were performed on limited number of studies (≤5) with less than 12 months of follow up duration.It is interesting that for some outcomes only one RCT was available, so the results seem unreliable.Thirdly, our results showed high evidence of statistical heterogeneity between the studies in some analyses which weakens the clinical certainty of the results (56, 57).Unfortunately, a low number of primary RCTs made it impossible to conduct subgroup analyses, so we were unable to find sources of heterogeneity between studies (n < 10).Fourthly, the effects of an intervention on selected outcomes are not solely based on statistical significance but should also be judged on clinical relevance.For example, the results of the current umbrella review showed inconsistent findings regarding the potential effects of pro-, pre-, and synbiotics supplementation on WC in patients with PCOS.Accordingly, probiotic supplementation slightly, but not clinically important, increased WC (0.86 cm) compared to the control group.In contrast, synbiotics and prebiotic supplementation decreased WC by nearly −2.7 and − 5.10 cm, respectively.Of course, these findings with low-quality evidence were obtained from data from only two trials for probiotics and one trial for synbiotics and prebiotics.Also, possible explanations for this inconsistency might be the short duration of the interventions.It is recommended that extend the treatment period for central obesity beyond 12 weeks (70, 71).Fifthly, it is imperative to consider strain-specific efficacy when using probiotics or symbiotics in the treatment or prevention of disease.The efficacy of potential probiotic strains varies according to experimental studies (72).As a result, it is important to determine whether the microbes can survive from ingestion to delivery to the target organ, whether the microbes are capable of interfering with pathogenesis (usually using animal models of disease), and whether they can be sustained from ingestion to administration (73).Interestingly, among 127 studied Lactobacillus strains, only 3% were found to be capable of being used as probiotics due to their ability to survive in the target organ and to withstand bile and stomach acidity (74).In addition, over 170 Lactobacillus species were examined in depth, revealing significant differences in resistance to antibiotics and probiotic potential (75   different factors could explain why some strains are effective in some types of diseases but are not effective in others.However, a direct comparison of different strains is relatively uncommon, and multiple trials for the same strain or mixture are not common for the same disease.Strain-specificity can be accounted for by including only probiotics belonging to the same strain in meta-analyses.Another strategy is conducting subgroup analyses with the same probiotic strains within each sub-group.The results of previous research showed that not all probiotic strains are as effective as originally believed based on subgroup analyses and re-analysis of the data (76)(77)(78).This critical point was not taken into account by any of the meta-analyses that included in this umbrella review.Our review on the primary included RCTs also showed that all of those studies intervened by mixture of probiotic strains.Among them, two trials intervened by symbiotic formulas with the same probiotic and prebiotic mixture (50,52) and two by capsules with the same probiotic mixture (45,48) while others contained different strains of probiotics.Accordingly, due to the lack of included primary studies, we were unable to perform subgroup analyses to cover this important note in detail.
The main mechanisms behind these beneficial effects of pro-, pre-, and synbiotics on PCOS-related outcomes are still unclear.However, one possible explanation may be due to the effects of these compounds on short-chain fatty acids (SCFAs), the main by-products of fermentation in the intestinal lumen.The production of SCFAs has been shown to influence intestinal mucosal integrity, resulting in reduced inflammation, microbial endotoxins, and insulin resistance.In addition, the SCFAs play a role in the regulation of food intake and blood glucose homeostasis through the regulation of the secretion of gut peptides such as peptide YY and glucagon-like peptide-1 (79).Moreover, it has been suggested that the SCFAs inhibit the activation of the rate-limiting enzyme in the cholesterol production pathway, hydroxymethylglutaryl-CoA reductase (HMG-CoA reductase), which leads to lower cholesterol metabolism and better lipid metabolism (80).Regarding sex hormones and hirsutism score, it has been found that probiotics or synbiotic supplements increase mucin formation, enhance bowel function, and reduce the quantity of gram-negative (inappropriate) bacteria in the colon.These modifications lessen the transmission of lipopolysaccharides (LPS) along the mucous wall and metabolic endotoxemia, which can ultimately result in improvements in insulin receptor function, lower levels of insulin, and increased levels of normal ovarian function, which in turn reduce the production of androgens such as DHEA, FAI, and testosterone (81, 82).As well, a limited number of RCTs with a short duration (less than 12 weeks) make it impossible to draw any conclusions regarding the impact of pro-pre-and synbiotic supplementation on PCOS-related outcomes, which adds to the importance of further studies in this area.
Our study had some strengths.This is the first study evaluating the effects of pro-, pre-, and synbiotic supplementation on several outcomes in patients with PCOS.To conduct this review, we selected the largest meta-analyses for each outcome, excluded RCTs without inclusion criteria, and recalculated effect sizes for each outcome, whenever possible.In addition, the certainty of the evidence was assessed using the GRADE tool.As a valid and acceptable tool, it helps the findings of systematic reviews to be more elucidative and informative.Accordingly, our review showed that, in most cases, the results of the meta-analyses were accompanied by small effect sizes and low or very low certainty of the evidence.
Our study has some limitations that should be considered.First, since the primary studies were limited to Iran and Egypt, these  Frontiers in Nutrition 14 frontiersin.orgfindings seem to have limited generalizability.Second, the number of studies for each outcome was limited and only one study has been conducted on the effects of prebiotics on PCOS-related outcomes.Third, the validity of our findings is impacted by considerable heterogeneity in some pooled results.Of course, we were unable to perform subgroup analyses to detect potential sources of heterogeneity because there were less than 10 trials available for each analysis.Forth, different probiotic and synbiotic supplementation across trials and the pooling of their effects added uncertainty to the interpretation of specific findings to each outcome.For example, although in the pooled data analysis probiotic supplementation improved FGC levels, synbiotic supplementation did not show any significant result.Fifth, the included meta-analyses did not obtain data from unpublished information, which may lead to publication bias.Sixth, it is impossible to fully control the confounding effects of other components of the diet via statistical methods, therefore, the effects of a pro-prebiotic and synbiotic supplementation may be partially mediated by other diet components.Seventh, the results of this study may be also cofounded by other PCOS-related lifestyle factors, such as body weight, age, and levels of physical activity.There were few primary studies, so we were unable to conduct subgroup analyses to take these factors into account.

Conclusion
In conclusion, the results of the present umbrella review suggests the beneficial effects of probiotics and synbiotics supplementation on the HOMA-IR index.However, the results originated from pooled data of the low number of RCTs with a maximum duration of 12 weeks.Also, we could not find a conclusive finding for other outcomes because of some important limitations such as small sample sizes in primary trials, small pooled effect sizes, and low or very low certainty in the evidence.Therefore, further well-designed RCTs with the following criteria might help to confirm or reject our findings in patients with PCOS: studies with different races and larger sample sizes; comparing the effects of different types of pro-, pre-, and synbiotic supplements on specific outcomes, RCTs with longer periods and larger sample sizes to assess and compare the effects of different dose of supplements, reporting all potential side effects following probiotics supplementation, and comparing the effects of different probiotics, prebiotics, and synbiotics to the promotion of evidence about the effects of these different interventions.
Our review generated several key messages for clinicians and patients, notably those who are eager for an adjuvant approach to the treatment of PCOS.Even though there are a variety of pathways that support the advantages of pro/pre and synbiotic supplementation in women with PCOS, it is critical to highlight that the magnitude of the effect was not clinically important, and the certainty of the evidence was low and very low.It is critical to highlight that there is insufficient data to support their obvious and long-term clinical effects.

FIGURE 1
FIGURE 1Literature search and review flow diagram for selection of umbrella review meta-analyses.

TABLE 1
General characteristics of the published meta-analyses investigating the effects of pro-pre/synbiotic supplementation in patients with polycystic ovary syndrome.

Table 4 )
. Detailed GRADE scores for each outcome are shown in Supplementary Table7.

TABLE 2
Characteristics of eligible primary studies on the effects of pro-pre/synbiotic supplementation in patients with polycystic ovary syndrome.

TABLE 3
The effects of probiotic supplementation in women with PCOS.

TABLE 5
The effects of prebiotic supplementation in women with PCOS.