Background: Polycystic ovary syndrome (PCOS) is a prevalent endocrine disorder among reproductive-aged women, one of the leading causes of anovulatory infertility. Women affected by PCOS consistently experience increased risk of adverse reproductive and pregnancy outcomes despite the advancement of assisted reproductive technologies (ART). 

Aim: The aim of this systematic review was to provide an overview of the recent evidence on the impact of PCOS on fertility and maternal-neonatal outcomes. 

Methods: We conducted a systematic review of our findings, using PubMed, Scopus, Web of Science as well as the Cochrane Library following the PRISMA 2020 standards. This search was conducted after incorporating randomized controlled trials, cohort studies and meta-analyses of reproductive health outcomes, primarily in women with PCOS (Rotterdam or NIH criteria). The outcomes measured were the rate of ovulation, time to conception, rate of pregnancy and live birth rate, miscarriage, gestational diabetes mellitus (GDM), preeclampsia and neonatal complications. The risk of bias was assessed by Cochrane RoB and ROBINS-I tools. 

Results: A total of over 92,000 participants were included in thirty-three studies. Women with PCOS have markedly decreased ovulation and pregnancy rates, longer time to conception, greater risks of miscarriage (OR 1.70), GDM (OR 1.51), and preeclampsia (OR 2.12). Neonatal outcomes included an increase in low birth weight and NICU admissions. Subgroup analyses showed that lean PCOS phenotypes had a greater response to ART protocols than obese ones. 

Conclusion: PCOS adversely affects reproductive outcomes, with increased maternal and neonatal risks. Some treatment strategies based on phenotypic and metabolic traits can, at least in part, improve fertility and pregnancy outcomes. More research is needed to achieve optimal ART protocols and reduce negative outcomes.

Polycystic ovary syndrome (PCOS) is a complex endocrine condition that affects some 6–20% of reproductive age women worldwide, determined by diagnostic criteria and population studied [1]. PCOS is most often associated with hyperandrogenism, ovulatory dysfunction, and polycystic ovarian morphology, and it is a leading cause of anovulatory infertility [2]. Nonetheless, up to 70% of affected women remain undiagnosed, leading to delayed treatment and increased risks for reproductive and metabolic complications [3].

PCOS pathophysiology is highly mediated by an interaction of genetic, hormonal, and environmental factors. Insulin resistance and compensatory hyperinsulinemia worsen hyperandrogenism, impairing folliculogenesis and ovulation [4]. Obesity, commonly linked to PCOS, exacerbates these adverse effects and leads to a vicious cycle that impairs reproductive function [5]. Increases in luteinizing hormone (LH) relative to follicle stimulating hormone (FSH) are more frequent with PCOS and play a role in abnormal ovarian steroidogenesis [6].

Clinically, PCOS frequently manifests with menstrual irregularities, hirsutism, acne, and infertility [7]. The syndrome is also associated with higher risk of miscarriage, gestational diabetes mellitus (GDM), preeclampsia (PE), and adverse neonatal outcomes including low birth weight (LBW) and preterm birth [8][9]. Such complications are especially prominent in women using assisted reproductive technologies (ART), as PCOS has been found to be associated with ovarian hyperstimulation syndrome (OHSS) and poor endometrial receptivity [10][11].

Clomiphene citrate, letrozole, and metformin—pharmacological treatments to induce ovulation—are widely used. Letrozole (an aromatase inhibitor) exhibits better effect in clomiphene resistant patients [12]. Metformin enhances insulin sensitivity and ovulatory function, particularly in obese and insulin resistant models [13]; thus, lifestyle changes—particularly weight loss and diet—remain the mainstay of PCOS therapy and have demonstrated significant reproductive improvement [14].

Even with availability of multiple treatment techniques, reproductive efficacy is variable in PCOS and reproductive outcome is inconsistent. ART protocols must be individualized to minimize risk and maximize success. A growing body of evidence shows that phenotype specific treatments can increase fertility and decrease complications [15]. Furthermore, the role of follicular fluid biomarkers and endometrial receptivity in predicting ART success is developing as well [16].

Due to the wide heterogeneity of PCOS and its effect on fertility, a full body synthesis of the available evidence is necessary. Previous literature reviews have examined only narrow interventions or outcomes and have not stratified by phenotype or mode of intervention. This review systematically evaluates reproductive outcomes—including ovulation, pregnancy, live birth, miscarriage, and time to conception—in women with PCOS across interventional and observational studies. This review combines subgroup analyses with risk stratification to guide clinical decision making and future research.

To provide methodological rigor, transparency and reproducibility, this systematic review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020). The protocol detailing research objectives, inclusion criteria, search strategy, data extraction, and planned analysis was prepared before the initiation of the studies and registered under the registration number CRD420251162018 in the International Prospective Register of Systematic Reviews (PROSPERO). Registration facilitated prospective documentation of review methods and reduced the potential for selective reporting.

Eligibility Criteria

The study selection adopted structure based on the PICOS model. Women aged 18–45 years diagnosed with polycystic ovary syndrome (PCOS) according to Rotterdam, NIH, or AE PCOS criteria were included in the population of interest. Compatible modes of study included randomized controlled trials (RCTs), cohort studies, case–control studies and systematic reviews or meta-analyses with reproductive outcomes in PCOS. The intervention types consisted of ovulation induction agents (clomiphene citrate, letrozole, metformin), lifestyle modification programmes, with ART (in vitro fertilization [IVF], intracytoplasmic sperm injection [ICSI]). Comparators utilized placebo, no treatment, or alternative treatment protocols. Primary outcomes were ovulation rate, clinical and biochemical pregnancy rates, live birth rate, miscarriage rate, and time to conception. 

Secondary outcomes were:

•  Gestational diabetes mellitus (GDM)

•  Preeclampsia

•  Neonatal complications

We excluded studies where they were:

•  Case reports

•  Editorials

•  Conference abstracts

•  Animal studies

•  Non-indexed literature

or where clear reproductive outcomes were not established.

Search Strategy

A systematic and extensive search of four main electronic databases – PubMed, Scopus, Web of Science and Cochrane Library has been performed. The search strategy included both Medical Subject Headings (MeSH) and free text terms (e.g., “Polycystic Ovary Syndrome,” “PCOS,” “reproductive outcomes,” “ovulation induction,” “pregnancy rate,” “live birth,” “miscarriage,” and “assisted reproduction”). To optimize the search and sensitivity, Boolean operators (AND, OR) were applied. Filters were used to limit results specifically to human studies in English published within the last ten years to ascertain the contemporary relevance at hand. To broaden our coverage as well, all reference lists of the participating studies and relevant systematic screened from the review databases and further eligible studies were searched in the reference lists to increase the coverage of all involved studies.

Study Selection

All retrieved records were imported into Rayyan QCRI, which is a web-based platform for blinded and independent screening. Reviewing titles and abstracts for relevance, two reviewers independently selected studies. Where eligibility was unclear or studies that met initial screening criteria, full text articles were obtained. Any discrepancies between reviewers were resolved in discussion or through consultation with a third reviewer. The study selection process was documented via a PRISMA flow diagram that recorded the number of records identified, screened, excluded and included.

Data Extraction

Extraction was achieved with a standardized and piloted Excel sheet. Included data was author, year, country of publication, study design, sample size, PCOS diagnosis criteria, as well as patient-specific intervention and comparator data, results reproductive outcome demographic information (age, BMI, phenotype), duration of follow-up, as well as statistical analysis methods. Other variables like sources of funding and conflicts of interest reported were also documented. Two reviewers independently extracted data; differences were resolved by consensus.

Quality Assessment

Reviews were manually 

Validated tools were utilized to assess the methodological quality and risk of bias of included studies. The Cochrane Risk of Bias 2.0 (RoB 2.0) tool was used to assess domains for randomized controlled trials including randomization process, deviations from intended interventions, missing outcome data, outcome measurement and selective reporting. Observational studies were rated with the Risk of Bias in Nonrandomized Studies of Interventions (ROBINS-I) tool, in which confounding, participant selection, intervention classification, deviations from intended interventions, missing data, outcome measurement, and reporting bias were evaluated. Three risk factors (low, moderate, or high risk of bias) of the studies were included. Bias distribution in the studies was represented graphically with summary tables and visual plots.

Integration of datasets and analysis of statistics

Narrative synthesis was performed for studies with heterogeneous designs or outcomes. Wherever homogeneity was available, meta-analysis was conducted by Review Manager (RevMan) software. Analysis of dichotomous outcomes was performed using pooled odds ratios (ORs) with 95% confidence intervals while continuous outcome analyses of weighted mean differences were made. Statistical heterogeneity (I², > 50% as significant) was considered. We performed predefined subgroup analyses based on PCOS phenotype, BMI category, types of intervention, and geographic region. Sensitivity analyses were conducted to assess the robustness of pooled estimates by excluding studies with a high risk for bias and/or extreme effect sizes.

Evaluation of Publication Bias and Evidence Quality

Publication bias was screened via visual funnel plots and Egger's regression test when applicable. The overall certainty of evidence for each outcome was judged against Grading of Recommendations Assessment, Development and Evaluation (GRADE) standards, taking into account risk of bias, inconsistency, indirectness, imprecision and publication bias. The outcomes were given high, moderate, low or very low certainty.

33 studies met inclusion criteria (randomized controlled and cohort studies, systematic reviews) comprising populations in Asia, Europe, and North America (Table 1). The participants are 45 to >3,000 participants and generally the population was diagnosed by Rotterdam criteria and interventions included lifestyle change, pharmacologic ovulation induction and/or assisted reproductive technologies (ART) of IVF and ICSI. 

Study Quality and Bias Evaluation

(Table 2) showed that risk of bias was low across the randomized domains, with five studies reporting overall low risk. Observational studies often mask allocation concealment and blinding. Finally, one study was deemed high risk because of incomplete reporting and suboptimal randomization procedures. The vast majority were carried out with methodological rigor, which allowed for credible synthesis.

Overall Reproductive Outcomes 

as Table 3 illustrates inter-treatment effectiveness comparison. Letrozole had the highest incidence of ovulation (OR 1.82, 95% CI 1.45–2.29) and pregnancy (OR 1.67, 95% CI 1.32–2.11) rates of all non-ART treatments. Metformin and lifestyle interventions exhibited some degree of success. ART procedures were associated with significantly elevated pregnancy (OR 1.89, CI 1.53–2.33) and live birth rates (OR 1.76, CI 1.42–2.19), but at the same time increased miscarriage risk (OR 1.31, CI 1.02–1.68). Heterogeneity in different outcomes was moderate (I² = 38–55%), suggesting moderate effects. 

Subgroup Analysis by PCOS Phenotype and BMI 

Table 4 showed distinct trends of subgroup analysis by PCOS phenotype and BMI. Phenotype A (classic PCOS) appeared to have the greatest risk of miscarriage (OR 1.82) and gestational diabetes (OR 1.51), whereas Phenotype D had better live birth rates (Table 4). The stratification by BMI indicated that lean PCOS patients had superior ART related outcomes such as more ovulation and birth rates, and lower miscarriage rates. Obese patients showed an increased risk for gestational diabetes (OR 1.72) and preeclampsia (OR 2.15) suggesting metabolic burden significantly contributes to reproductive prognosis. These findings endorse phenotypic and BMI-based therapy approaches in terms of fertility outcomes. 

Evidence Quality Assessment 

The evidence quality assessment summary is shown in Table 5. Results like ovulation and pregnancy rates were supported by moderate quality evidence with low bias risk and acceptable consistency. For live birth and miscarriage outcomes, which were low due to heterogeneity and imprecision from effect estimates. Publication bias was assessed in a funnel plot (Figure 3), with mild asymmetry among smaller studies indicating selective reporting could affect pooled outcomes. Egger’s regression intercept test was only marginally significant and cautious interpretations were indicated for the respective analyses. 

Biochemical Insights from Follicular Fluid Analysis 

as Table 6 integrates molecular data from follicular fluid analyses, suggesting that some reproductive success biomarkers may play important roles in the female reproductive tract during PCOS. Markers of oxidative stress, including 8-Isoprostane, were elevated and associated with higher miscarriage rates. Low antioxidant capability (e.g., TAC levels) was associated with poor oocyte quality. Lipidomic changes such as higher ceramide and free fatty acids were linked to decreased embryo viability and fertilization results. The inflammatory markers (such as TNF-α, IL-6), were higher in the PCOS cohorts, which hampered blastocystogenesis. Hormonal biomarkers, including AMH and estradiol had the paradoxical trend of elevation, which suggested a follicular arrest in the presence of high follicle counts. Placental growth factor (PlGF) was found to be an early marker predicting ovarian response in ART cycles. 

Visual Summaries 

The flow of PRISMA studies selection (Figure 1) can be summarized visually including study selection of 33 studies, where 26 studies (Table 1) and 914 cases included and included by inclusion of meta-analysis. Forest plots in Figure 2 represent pooled effect sizes over major reproductive outcomes it summarizes pooled effect estimates derived from meta-analysis studies comparing reproductive outcomes among women with PCOS and controls. The forest plot shows substantial increases in gestational diabetes (OR 1.51), preeclampsia (OR 2.12), and low birth weight (OR 1.29), suggesting moderate heterogeneity across studies (I² = 48%). This agreement underscores increased perinatal risks of PCOS that underscore the importance of personalized antenatal protocols and metabolic risk screening. 

As shown in Figure 3, the funnel plot that measures publication bias. The effect sizes scatter pattern versus standard errors reveals some subtle asymmetry with the small studies reporting good results. Most are in standardized confidence interval, but a number of outliers result from selective reporting or small-study effects. Egger’s regression intercept test presents near significant, cautioning caution as well as transparency as for future studies of PCOS. A stratified forest plot comparing reproductive outcome in women with PCOS for different phenotypes (A–D) and BMI (lean, overweight, obese) is presented in Figure 4. 

The results show that phenotype A, (classic PCOS), is the type with the highest miscarriage and gestational complications, whereas phenotype D demonstrates a less severe prognosis. Lean individuals with PCOS demonstrate early ovulation and live birth after ART, and obese populations show increased likelihood of developing gestational diabetes (OR 1.72) and preeclampsia (OR 2.15). Effect sizes between subgroups are moderate to high heterogeneity, highlighting the effect of endocrine profile and body composition on fertility interventions. This number highlights the significance of phenotype- and weight-mediated-guided treatment regimens for achieving optimal reproductive attainment in women with PCOS.

Figure 1: PRISMA Flow Diagram

Legend: PRISMA diagram summarizing identification, screening, eligibility assessment, and final inclusion of studies in the review.

Figure 2: Forest Plot of Meta-analysis

Legend: Forest plot presenting pooled odds ratios for reproductive outcomes across interventions among    women with PCOS.

Figure 3: Funnel Plot for Publication Bias

Legend: Funnel plot evaluating publication bias and small‑study effects across included studies in the meta‑analysis

Figure 4: Subgroup Forest Plot by Phenotype and BMI

Legend: Subgroup forest plot comparing reproductive outcomes across PCOS phenotypes and BMI categories in included studies.

This systematic review and meta-analysis synthesized data from 33 studies evaluating reproductive outcomes in women with PCOS, with emphasis on BMI/phenotype stratification, ART response, and biochemical information obtained from follicular fluid (FF). The evidence demonstrates distinct reproductive risk profiles based on endocrine subtype and metabolic status. Women with PCOS undergoing ART experienced higher risks of gestational diabetes mellitus (GDM), preeclampsia (PE), miscarriage, and low birth weight (LBW) [20][21]. These complications were especially pronounced among classic phenotype A and obese individuals, reflecting the interplay between hyperandrogenism, insulin resistance, and systemic inflammation [22].

Subgroup analysis showed that lean PCOS phenotypes (BMI <25>

Phenotypic variation also influenced pregnancy outcomes. Phenotype D (ovulatory dysfunction/polycystic ovaries) demonstrated improved outcomes when lifestyle factors were optimized, whereas phenotype A showed increased miscarriage risk and poorer ART response, likely due to combined hormonal and inflammatory disturbances [26]. Publication bias—identified through funnel plot asymmetry and Egger’s test—suggested overrepresentation of small positive studies. GRADE assessment indicated moderate quality evidence for ovulation and pregnancy rates, but low-quality evidence for live birth and miscarriage due to heterogeneity and variability [27].

Biochemical characterization of FF in PCOS revealed elevated oxidative stress markers (e.g., 8 Isoprostane) and reduced total antioxidant capacity (TAC), both associated with poor oocyte quality and increased miscarriage [28][29]. These redox disturbances impair mitochondrial function and granulosa cell dynamics, contributing to subfertility. Lipidomic analyses identified elevated ceramide (Cer 36:1;2) and free fatty acid (FFA C14:1) concentrations in PCOS FF, which correlated negatively with embryo quality and fertilization success [30]. Reduced lysophosphatidylglycerol (LPG 18:0) suggested compromised membrane fluidity and impaired oocyte maturation [31].

In obese PCOS phenotypes, inflammatory markers such as TNF α, IL 6, and CRP were persistently elevated, indicating chronic low-grade inflammation associated with impaired blastocyst formation and implantation failure [32]. These cytokines disrupt granulosa cell signaling and steroidogenesis. Hormonal FF markers including AMH and estradiol were paradoxically increased, reflecting high follicular count but follicular arrest. Placental Growth Factor (PlGF) emerged as a potential indicator of ovarian responsiveness, with elevated levels predicting hyper responsiveness to gonadotropins [33]. Proteomic analyses in PCOS FF identified abnormalities in oxidative phosphorylation, MAPK signaling, and cell cycle checkpoint pathways, suggesting potential therapeutic targets to enhance reproductive performance [34].

Lifestyle interventions—particularly low glycemic diets and structured exercise—significantly improved insulin sensitivity and restored ovulatory cycles [35]. Emerging evidence supports the role of nutritional supplements such as inositol, omega 3 fatty acids, and chromium in ameliorating hormonal and metabolic abnormalities in PCOS. A recent network meta-analysis ranked these agents favorably for lipid modulation and insulin regulation [36].

Despite strong findings, limitations persist. Generalizability is restricted by heterogeneity in diagnostic criteria, phenotype classification, and intervention design. Standardizing FF biomarker thresholds and incorporating phenotype stratified outcomes in future trials would strengthen the evidence base. This review underscores the importance of tailoring reproductive management to the metabolic and endocrine profile of PCOS. Stratified treatment approaches guided by biochemical and phenotypic markers may optimize ART outcomes and reduce pregnancy complications. FF biomarkers offer promising avenues for personalized fertility protocols, particularly for high risk PCOS subgroups.

Strength

Strengths of this review include detailed phenotype and BMI stratification, integration of mechanistic FF biomarkers, and synthesis across 33 studies with moderate to low bias. Global representation enhances external validity. GRADE based assessment provided transparency regarding evidence quality. 

Limitations

Limitations include study heterogeneity, incomplete phenotype stratification, short follow up durations, publication bias, and underrepresentation of lean PCOS phenotypes, highlighting the need for standardized, longitudinal research.

This review finds that phenotypic presentation and BMI are associated with reproductive outcomes in women with PCOS. Classic phenotype A and obesity were consistently correlated with lower ART success and increased gestational complications, while lean phenotypes demonstrated better fertility outcomes. Follicular fluid biomarkers highlighted the underlying oxidative, lipidomic, and inflammatory dysregulation underlying subfertility. The integration of personalized reproductive regimes with a focus on metabolic and endocrine diversity for PCOS was supported by these findings. 

Recommendations

In patients with PCOS, care based on phenotype and BMI are paramount to optimize reproductive results for optimal treatment. Follicular fluid biomarkers improve prognostication and inform ART planning. A common diagnosis framework is also needed for consistent stratification and reporting. Long term follow-up studies of maternal and neonatal follow-up is essential to measure the larger effects. And finally, predictive models driven by AI can help make fertility care better and more personalized — from accurate risk assessment to personalized reproductive intervention.

Acknowledgements

The authors would like to extend their heartfelt thanks to Dr. Rayia Abdelwahid for her profound clinical advice, well-reasoned scholarly input, and dedication during the preparation process of this writing. 

Author Contributions

Awadalla Abdelwahid conceptualized the study and performed the statistical analysis. Awadalla Abdelwahid, Fathelrahman Elrasheed, and Hajar Suliman conducted screening, data extraction, quality assessment, and drafted the manuscript. Bashir Abdeen, Ahazeej Gurashi, and Nisrin Magboul Elfadel contributed to the methodology, search strategy, and manuscript review. All authors interpreted the findings, approved the final manuscript, and accepted responsibility for the integrity of the work.

Funding

Neither governmental, commercial, nor nonprofit funding bodies provided financial support for this study. 

Ethical Approval

Since this systematic review was based solely on research published prior to this study and had neither direct human participants nor identifiable data, there was no requirement to obtain formal ethical approval. All studies included had ethics clearance from their institutions. To ensure methodological transparency, the review protocol was prospectively registered in PROSPERO (CRD420251162018). 

Conflict of Interest

The authors declare they have no conflicts of interest in the conduct, analysis, or publication of this research. No personal, financial, professional, or other ties to a client shaped and influenced the findings in this manuscript.

Data Availability

All the data used for this review comes from existing publicly published studies indexed in databases such as PubMed, Scopus, Web of Science, and the Cochrane Library. It generated no primary new data. Such extracted datasets, summarized tables, and analytical materials can be obtained from the respective author on a reasonable request.

List of Abbreviations

PCOS         Polycystic Ovary Syndrome

ART           Assisted Reproductive Technologies

IVF            In Vitro Fertilization

ICSI           Intracytoplasmic Sperm Injection

BMI           Body Mass Index

FF               Follicular Fluid

GDM          Gestational Diabetes Mellitus

PE              Preeclampsia

AMH          Anti-Müllerian Hormone

TAC           Total Antioxidant Capacity

OR             Odds Ratio

CI               Confidence Interval

PlGF           Placental Growth Factor

LPG           Lysophosphatidylglycerol

FFA            Free Fatty Acids

CRP           C-Reactive Protein

TNF-α        Tumor Necrosis Factor-alpha

IL-6            Interleukin-6

RCT           Randomized Controlled Trial

PRISMA   Preferred Reporting Items for Systematic Reviews and Meta-Analyses

AE-PCOS  Androgen Excess Polycystic Ovary Syndrome Society criteria