Effect of menstrual cups on vaginal health, reproductive tract and urogenital infections: a systematic review and meta-analysis Authors: Anna Maria van Eijk, Alexandra Quinn-Savory, Garazi Zulaika, Alys Skinner, Linda Mason, Stefan J. Green, Penelope A. Phillips–Howard, Supriya D. Mehta Published: 2026 ABSTRACT SummaryBackgroundThe menstrual cup is a device inserted into the vagina during menstruation to collect blood. Evidence of its effect on the vaginal microbiome and associations with laboratory-confirmed infections is limited. We compiled existing information to assess associations between menstrual cups and reproductive tract infections (RTIs), sexually transmitted infections (STIs), and the vaginal microbiome.MethodsWe searched four databases (PubMed, CINAHL, Global Health, and Scopus, from inception to January 23, 2026) for trials or observational studies reporting on menstrual cup use and laboratory-confirmed RTIs, STIs and the vaginal microbiome. Study quality was assessed using the Cochrane Collaboration tool for trials and the Joanna Briggs Institute Critical Appraisal Checklist for Cohort Studies. Outcomes of interest between cup-users and users of alternative menstrual products (non-cup users) were compared using prevalence ratios (PR). Meta-analyses were conducted using fixed-effects models to generate pooled estimates. PROSPERO registration: CRD42024559595.FindingsEleven studies involving 10,268 participants were included. Only three studies were considered of good quality. In two randomised controlled studies in Kenya, menstrual cups reduced the risk of STIs relative to non-cup users (adjusted PR [aPR] 0.74, 95% CI 0.60–0.91, p = 0.037, n = 695, I2 = 0.0%). The association with bacterial vaginosis was 0.81 (aPR, 0.63–1.04, p = 0.0963, n = 695, I2 = 0.0%). When combining observational and randomised studies, a protective effect on bacterial vaginosis persisted (PR 0.79, 95% CI 0.69–0.90, p = 0.0004, five studies, n = 2103, I2 = 0.0%, 3 continents). One trial showed an association between menstrual cups and HSV-2 of 0.71 (aPR, 95% CI 0.50–1.01, p = 0.057, n = 1451). Optimal vaginal community state type-1 (CST-I) was more likely among cup users compared to non-cup users (PR 1.18, 1.10–1.27, p < 0.0001, n = 2639, 5 studies in Kenya and Europe). No increased infection risks were noted among menstrual cup users across the infections evaluated. Main limitations included the low number and quality of existing studies.InterpretationMenstrual cups reduced the risk of STIs in Kenya. For bacterial vaginosis, the association was consistent with a protective effect and supportive for a healthy vaginal microbiome composition in studies from different geographies. These data strengthen evidence on the value of menstrual cups as a global multipurpose menstrual product solution.FundingUK-10.13039/501100000265Medical Research Council, 10.13039/501100009187Medical Research Foundation, USA-10.13039/100000002National Institutes of Health; 10.13039/100000071National Institute of Child Health and Human Development, and the Gates Foundation. Research in contextEvidence before this studyPrevious studies have shown that menstrual cups are cost-effective, durable, safe and accepted among diverse populations for management of menstruation. As menstrual product and material access expands, understanding the risks and benefits of menstrual cups for sexual and reproductive health outcomes is important for future users, programme managers, and policy makers. We searched PubMed, CINAHL, Global Health, and Scopus from inception to January 23, 2026, using the search terms ((“Menstrual Cup” OR “Vaginal Cup”) AND Infections AND review). We included reviews with information on menstrual cup use and infections and identified four studies of interest. In a 2019 review of multiple studies, 73% of 1144 participants were satisfied with the menstrual cup and wished to continue use. Two reviews reported on one case and five cases of toxic shock syndrome in association with menstrual cup use. One review listed infection risks and reported no increased risk of reproductive tract or systemic infection risk associated with menstrual cup use compared with other menstrual products among European, North American, and African women and girls. One review published in 2025 examined the relationship between menstrual products (any) and reproductive tract infections by self-report or laboratory test. However, results were only described in the text, and no meta-analyses were conducted. Knowledge and methodologies for assessing vaginal health have advanced, and new studies have examined the impact of menstrual cups on reproductive tract infections, including sexually transmitted infections, and the vaginal microbiome.Added value of this studyThis systematic review and meta-analysis focuses on laboratory-confirmed reproductive tract and sexually transmitted infections and the vaginal microbiome, and includes data from 11 studies, of which eight were published after 2019. This review summarises data on the association between these infections and cup use, providing pooled estimates where applicable. Compared to non-cup users, cup-users had lower proportions of bacterial vaginosis across study populations from the USA, Denmark, Tanzania, and Kenya, and reductions in sexually transmitted infections from studies in Kenya. Higher proportions of a vaginal bacterial community type dominated by Lactobacillus crispatus were found among cup-users compared to non-cup users, from populations in three countries in Europe and one country in Africa. L. crispatus is known to confer protection against vaginal dysbiosis and sexually transmitted infections. These findings strengthen evidence for policymakers and health programmes, indicating that menstrual cups can serve as a multipurpose tool that support menstrual care while promoting a healthy vaginal microbiome and improving sexual and reproductive health.Implications of all the available evidencePromoting reproductive tract health through menstrual cup use offers a low-cost, scalable and sustainable solution to reducing the burden of preventable gynaecological conditions, while supporting menstrual hygiene. These benefits are particularly significant in low-resource settings, where limited access to healthcare heightens the risk of adverse outcomes associated with vaginal dysbiosis and infection. However, there was substantial uncertainty in evidence stemming largely from the need for more randomised trials in broader populations, with larger sample sizes. Future research should focus on strengthening the evidence and elucidating the mechanisms by which the menstrual cup improves the vaginal microbiome. Research in context Evidence before this study Previous studies have shown that menstrual cups are cost-effective, durable, safe and accepted among diverse populations for management of menstruation. As menstrual product and material access expands, understanding the risks and benefits of menstrual cups for sexual and reproductive health outcomes is important for future users, programme managers, and policy makers. We searched PubMed, CINAHL, Global Health, and Scopus from inception to January 23, 2026, using the search terms ((“Menstrual Cup” OR “Vaginal Cup”) AND Infections AND review). We included reviews with information on menstrual cup use and infections and identified four studies of interest. In a 2019 review of multiple studies, 73% of 1144 participants were satisfied with the menstrual cup and wished to continue use. Two reviews reported on one case and five cases of toxic shock syndrome in association with menstrual cup use. One review listed infection risks and reported no increased risk of reproductive tract or systemic infection risk associated with menstrual cup use compared with other menstrual products among European, North American, and African women and girls. One review published in 2025 examined the relationship between menstrual products (any) and reproductive tract infections by self-report or laboratory test. However, results were only described in the text, and no meta-analyses were conducted. Knowledge and methodologies for assessing vaginal health have advanced, and new studies have examined the impact of menstrual cups on reproductive tract infections, including sexually transmitted infections, and the vaginal microbiome. Added value of this study This systematic review and meta-analysis focuses on laboratory-confirmed reproductive tract and sexually transmitted infections and the vaginal microbiome, and includes data from 11 studies, of which eight were published after 2019. This review summarises data on the association between these infections and cup use, providing pooled estimates where applicable. Compared to non-cup users, cup-users had lower proportions of bacterial vaginosis across study populations from the USA, Denmark, Tanzania, and Kenya, and reductions in sexually transmitted infections from studies in Kenya. Higher proportions of a vaginal bacterial community type dominated by Lactobacillus crispatus were found among cup-users compared to non-cup users, from populations in three countries in Europe and one country in Africa. L. crispatus is known to confer protection against vaginal dysbiosis and sexually transmitted infections. These findings strengthen evidence for policymakers and health programmes, indicating that menstrual cups can serve as a multipurpose tool that support menstrual care while promoting a healthy vaginal microbiome and improving sexual and reproductive health. Implications of all the available evidence Promoting reproductive tract health through menstrual cup use offers a low-cost, scalable and sustainable solution to reducing the burden of preventable gynaecological conditions, while supporting menstrual hygiene. These benefits are particularly significant in low-resource settings, where limited access to healthcare heightens the risk of adverse outcomes associated with vaginal dysbiosis and infection. However, there was substantial uncertainty in evidence stemming largely from the need for more randomised trials in broader populations, with larger sample sizes. Future research should focus on strengthening the evidence and elucidating the mechanisms by which the menstrual cup improves the vaginal microbiome. Introduction Every month, almost 1.8 billion people menstruate worldwide, bleeding on average 65 days a year.1 Despite the need for effective, safe and affordable menstrual products, few solutions are available to manage menstruation.2 For products that do exist, limited awareness and/or availability, cost, prejudice, and safety concerns impede their uptake. In a review of 42 nationally-representative surveys in low- and middle-income countries, the use of reusable menstrual materials (not further specified in these surveys, but often cloths or rags) was estimated at 12%, and was associated with lower education levels, lower socio-economic status, and lower country gross domestic product.3 Non-reusable products such as disposable pads are dominant, and often cited as the most accessible and convenient to use.4,5 However, single-use menstrual products require continued purchase and are environmentally polluting.6 In impoverished settings people who menstruate report using a range of non-hygienic, makeshift materials, such as rags, cotton wool, or tissues.7,8 These materials can result in discomfort, shame, leakage, odour and elevated risk of urogenital infections, including bacterial vaginosis.9, 10, 11, 12, 13, 14 Menstrual cups are receptacles that are inserted into the vagina to collect menstrual flow. They can be used for up to 12 h prior to being emptied and reinserted. ‘Vaginal cups’ are bell-shaped and sit at the opening of the vaginal canal, while ‘cervical cups’ are disc-shaped and placed higher up against the cervix.15 Most menstrual cups are reusable and made of medical-grade silicone, rubber, latex or elastomer. Early studies conducted to explore the safety of menstrual cups reported no changes or abrasions to the vaginal wall or abnormalities in the vaginal flora.16 A systematic literature review and meta-analysis found minimal safety issues related to cup use and indicated that menstrual cup users had similar or less frequent blood leakage when compared to those using other methods of menstrual hygiene management.17 Menstrual cups are growing in popularity as a menstrual hygiene device, with individuals reporting they chose menstrual cups for economic, comfort and environment related reasons.4 A large digital cohort based in the United States (n = 11,455, time under study 2019–2024) reported 18% of participants used a menstrual cup.4 In low- and middle-income settings, where substantial barriers to achieving menstrual health and hygiene exist, emerging studies suggest menstrual cups can be an acceptable solution to managing menses.18 Currently, no systematic review or meta-analysis has been conducted to examine the effect of menstrual cups on reproductive tract infections, including bacterial vaginosis. This last condition represents a shift in the vaginal microbiome from one that is Lactobacillus dominant to one that is diverse and having anaerobic bacteria.19,20 Transient fluctuations in vaginal microbiota are observed throughout the menstrual cycle, and hormonal contraception and sexual activity can affect microbiota.21 Additionally, evidence demonstrates an association between poor menstrual hygiene, such as use of cloths, and non-optimal vaginal microbiome containing lower relative abundance of L. crispatus.12 Numerous studies have illustrated the effect of bacterial vaginosis and non-optimal vaginal microbiota on poor maternal outcomes, such as heightened risk of preterm birth,22,23 as well as increasing the risk of acquiring STIs and HIV.24, 25, 26 Should menstrual cups be found to reduce risk of infections and help in maintaining a Lactobacillus-dominated vaginal microbiome, it would be a multipurpose tool that improves sexual and reproductive health (SRH) and supports the management of menstruation. This systematic review and meta-analysis investigates the association between menstrual cups and reproductive tract infections including STIs, and the vaginal microbiome. Methods Eligibility This review was registered in PROSPERO (CRD42024559595) and followed the PRISMA guidelines. Studies were included if they were quantitative, had information on menstrual cup use, and included at least one biological or laboratory measure or a clinical syndromic diagnosis as an outcome. Eligible designs included before/after studies, cross-over or cohort designs, and randomised, pilot, quasi-randomised, or cluster-randomised designed studies or trials (henceforth grouped as ‘trials’) and cross-sectional studies. Studies reporting information with insufficient quantitative detail for forest plots were included in a table of available information. Qualitative studies, cup studies with no biological measures, laboratory-based studies where no human participants used cups, or studies where the menstrual cup was solely used to collect cervicovaginal fluids were ineligible. Information sources Sources searched for relevant literature included PubMed, CINAHL, Global Health, and Scopus databases, using the search terms listed in Appendix p2. Where available, reference lists of material of interest, conference abstracts and reports were checked. No time or language restriction was incorporated in the search, and databases were searched from their inception. Google Scholar, sorted chronologically, was searched last until there was no new material for 5 consecutive pages. The search was conducted on May 3, 2025, and an update on 23 January 2026 did not result in new material. Selection process All citations were imported into Excel and duplicates were removed. Two reviewers (AMvE and AQS) independently screened the abstracts and, where applicable, the full texts, and subsequently agreed on final study eligibility. Any disagreements were resolved by a third reviewer (GZ). A log of all excluded studies and reasons for exclusion was kept (Fig. 1). Data from studies with multiple publications were combined in one record entry to avoid duplication of individual data.Fig. 1Prisma flow diagram. ∗See Appendix p13. †The reports by Phillips–Howard and colleagues, 2016, and Juma and colleagues, 2017, had information from the same study. The publication of Mehta and colleagues, 2023, was a sub-study of the trial reported in Zulaika and colleagues, 2023, without overlap in infections determined and was considered a separate study. Prisma flow diagram. ∗See Appendix p13. †The reports by Phillips–Howard and colleagues, 2016, and Juma and colleagues, 2017, had information from the same study. The publication of Mehta and colleagues, 2023, was a sub-study of the trial reported in Zulaika and colleagues, 2023, without overlap in infections determined and was considered a separate study. Data collection process Two people (AMvE and AQS) independently extracted data from the included studies and conducted risk of bias assessment using the Cochrane Collaboration tool for trials. This form collects data on the study enabling a risk of bias assessment using seven categories, visualised using a risk-of-bias plot tool. Randomised trials could obtain a maximum of 7 points. A tool for cohort studies from the Joanna Briggs Institute was used for observational studies (cohorts and surveys, because surveys can be considered as retrospective cohorts for this analysis) with a maximum of 11 points.27 Any disagreements in bias assessment were resolved by a third reviewer (GZ). A score of 6 or 7 for trials, and 9–11 for observational studies was considered a good quality study. Data items extracted included first author, journal, year of publication, study design, study setting, type of cup used, and outcome measures by menstrual cup use, including method of laboratory assays or detection. We attempted to contact authors if there was insufficient clarity about the extracted data, or if data were not in an extractable format. Outcomes included: reproductive tract infections (as binary variables)–bacterial vaginosis, aerobic vaginitis and Candida albicans; sexually transmitted infections–Trichomonas vaginalis, Chlamydia trachomatis, Neisseria gonorrhoeae, herpes simplex virus-2 (HSV-2), human immunodeficiency virus (HIV), and human papillomavirus; vaginal colonisation of Staphylococcus aureus, or Streptococcus agalactiae; vaginal microflora—Lactobacillus spp; and other urogenital infections–female genital schistosomiasis, Escherichia coli and urinary tract infections. If information about infections was reported but not confirmed by laboratory tests these data were not included in the main analyses but were reported in the Appendix. Information on the vaginal microbiome was dichotomised as vaginal community state type 1 (CST-I, which is L. crispatus dominated) vs. other CST (CST-II with Lactobacillus jensenii dominance, CST-III with Lactobacillus iners dominance, CST-IV which is polymicrobial, and CST-V with Lactobacillus gasseri dominance).28 CST-I is considered optimal, being consistently associated with reduced risks of bacterial vaginosis, STIs, HIV, and adverse pregnancy outcomes.29 The relative abundance of L. crispatus (continuous variable as percentage) was also included as an outcome. Synthesis (Statistics) Study characteristics and method of recruitment were tabulated. Outcomes of interest between cup-users and non-cup users were compared using prevalence ratios. For the study by Mehta and colleagues, which reported odds ratios, prevalence ratios were calculated from the available dataset using the GLLAMM procedure in Stata (Generalised Linear and Latent Mixed Models by maximum likelihood with log link and binomial distribution, and repeated visits accounted for with random effect for the individual). Meta-analysis was conducted using Stata (version 17), with a fixed-effects model employed to estimate pooled prevalence ratios. The I-squared (I2) was used as an indicator of heterogeneity and quantifies the percentage of variation across studies that is due to heterogeneity rather than chance. Unadjusted prevalence ratios were presented in forest plots with pooled estimates where applicable, and adjusted prevalence ratios where available. Forest plots were additionally used as a tool to present available information without pooling results. Sources of heterogeneity would have been determined through subgroup analysis if there had been sufficient studies (ten or more).30 Sensitivity analysis was conducted using the quality assessment of the study (low-to-moderate vs. good quality study) and a leave-one-out analysis was conducted for outcomes with five or more studies available. Although the number of studies per outcome was low, the funnel plot for bacterial vaginosis was presented to allow assessment of publication bias. Additional information on changes to the protocol and the analyses can be found in the Appendix (p4). The certainty of evidence was evaluated post-hoc using the GRADE approach31; details on the method are available in the Appendix (p4). Ethics No ethical approval was sought for the conduct of this systematic review and meta-analysis, because no primary data collection involving human subjects took place. All data presented and analyses conducted are based on previously published research. Role of funders The funders had no role in the study design, data collection, analysis, interpretation, or writing of this manuscript. The corresponding author had full access to all the data in the study and together with the joint senior authors had final responsibility for the decision to submit for publication. Results Characteristics The database searches identified 85 items, resulting in 44 after deduplication, of which eleven studies, involving 10,268 participants described in 12 records, were eligible (Fig. 1, Table 1). The study by Mehta and colleagues was a sub-study of the trial reported by Zulaika and colleagues, but because of the different pathogens studied, was considered a separate study.33,34 The included studies were conducted across low-, middle- and high-income countries (Kenya, Tanzania, Brazil, France, Belgium, Denmark, and USA), and varied in designs, recruitment methods, interventions tested, and outcomes examined (Table 1).13,16,21,32, 33, 34, 35, 36, 37, 38, 39, 40 Two studies were cluster-randomised trials, one was a cluster-randomised pilot study, one study had a randomised cross-over design, and the remaining seven studies were observational. The age range varied from adolescent schoolgirls to women of reproductive age (range 14–55 years). Three studies were considered of good quality (Appendix p5). None of the observational studies reached maximum quality score; the inability to measure the exposure (the use of the menstrual cup) beyond a verbal report was a factor contributing to their low scores. Further details on included studies and laboratory tests can be found in the Appendix p7–p12.Table 1Studies contributing to this review.Author, yearCountryTime periodDesignSample size and age, sexual exposure baseline if cohortCup brand (type), sexual exposure baseline if trialComparison, sexual exposure baseline if trialFollow up and cup adherenceLoss to follow-up (%)Study qualityaCohort studies or surveys with quantitative information included in forest plotsNorth 201116USANRCohort: before-after study406 women, range 18–55 yrs.Sexually active: 86.6%Soft cup (cervix cup, disposable)Non-cup user: Pads (18%), tampons (31%), pads & tampons (51%)3 menstrual cyclesAdherence 75.9%24.1% (98/406)6 out of 11Phillips–Howard 2016,13 Juma 201732Kenya2012–2013Cluster-randomised pilot study751 girls, range 14–16 yrs (mean: 14.6, SD 0.7)MooncupEver had sex: 88.3%Non-cup user:Pads (40%) or “usual material”Ever had sex 84.0%12–14 months, median 10.9Adherence: 70.8%33.2% (249/751)6 out of 7Zulaika 202333Kenya2017–2021Cluster-randomised trial4137 girls (median: 17.1, IQR 16.3–18.0)MooncupSexually active: 28.2%Non-cup user: Pads (61%) or “usual material”Sexually active: 27.1%2–3 years, median 2.5Adherence: 79.3%12.4% (514/4137)5 out of 7Mehta 202334,bKenya2018–2021Cluster-randomised trial436 (median: 16.9)MooncupSexually active: 35.6%Non-cup user: Pads 72%, cloth 3%, pads and cloth 25%Sexually active: 31.4%30 monthsAdherence estimated at 79.3%3.2% (14/436)6 out of 7Tessandier 202335France2016–2020Retrospective cohort138 women, 18–25 years (intervention median: 23.0; control: 21.0)Menstrual cup 75% of the timeIntercourse last 2 weeks: 72.9%Tampon 75% of the time. Intercourse last 2 weeks: 71.0%24 months (assessment of menstrual item use)Unclear6 out of 11Sica 202236USANRIndividually randomised cross-over trial65 women, 18–55 years (mean 34.2, SD NA, median 33, range 19–49)Tampax cup and another cupBaselineTampon 92%Cup 8%Baseline and 2 cycles (1 cycle with each product)f15.4% (10/65)6 out of 7Graciolli 202337Brazil2019Prospective cohort47 participants, 18 yrs and older (mean: 24.0, SD 4.7).Sexually active: 91.5%MooncupDisposable pads3 cycles. Adherence 97.9%Unclear6 out of 11Femme International38,cTanzania2019–2020Prospective cohort79 women, 20–50 yrs, mean 33, SD NALunette cupsNon-cup user (materials not specified)Cup use 2–3 yearsfNot reported4 out of 11Haahr 202539Denmark2017–2022Survey in 4 IVF-clinics1153 IVF patients, 18–42 years (mean 32.0, SD 5)No brand reportedNon-cup user:Pads (46%), tampons (29%), pads & tampons (24%)NAf1153/1511 available (76.3%)d6 out of 11Lebeer 202321Belgium2020–2020Cross-sectional study3043 women, ≥18 yrs, median ≈29 yrsNo brand reportedNon-cup user:Pads, tampons, liners & combinations.Percentage NANAf4.5% (149/3345)e6 out of 11Cohort studies or cross-sectional studies with insufficient quantitative information but statements on infectionsKarnaky 196240USANRProspective cohortsCohort 1: 50Cohort 2: 97Age not reportedTassette (vaginal)Non cup user: pads, tampons. Percentage NA24 months?UnclearfUnclear1 out of 11IQR, interquartile range; IVF, in-vitro fertilization; NA, not available; Pads, sanitary towels; SD, standard deviation; Yrs, years.Adherence: verbal report of menstrual cup use. Not available for Mehta and colleagues (2023), information used from Zulaika 2023.aA tool from the Cochrane Collaboration was used for trials (maximum of 7 points) and from the Joanna Briggs Institute for observational studies (maximum of 11 points).bMehta and colleagues, 2023, was a sub-study of Zulaika and colleagues, 2023, including 6 schools and only two study arms (menstrual cups vs. control) and more intense follow up (6, 12, 18 and 30 months) without overlap in infections reported and was considered a separate study.cData from this study were extracted as our best estimates from a figure in the publication. A request for more information about this study did not result in the acquisition of better data.dThe question about menstrual product use was not started from the beginning of the study.e4682 women registered with questionnaire, and 3345 women sent a self-swab vaginal sample. However, 149 out of 3345 samples did not pass the quality test.fNo information on sexual activity of participants or about menstrual cup adherence in this study. Table 1Studies contributing to this review.Author, yearCountryTime periodDesignSample size and age, sexual exposure baseline if cohortCup brand (type), sexual exposure baseline if trialComparison, sexual exposure baseline if trialFollow up and cup adherenceLoss to follow-up (%)Study qualityaCohort studies or surveys with quantitative information included in forest plotsNorth 201116USANRCohort: before-after study406 women, range 18–55 yrs.Sexually active: 86.6%Soft cup (cervix cup, disposable)Non-cup user: Pads (18%), tampons (31%), pads & tampons (51%)3 menstrual cyclesAdherence 75.9%24.1% (98/406)6 out of 11Phillips–Howard 2016,13 Juma 201732Kenya2012–2013Cluster-randomised pilot study751 girls, range 14–16 yrs (mean: 14.6, SD 0.7)MooncupEver had sex: 88.3%Non-cup user:Pads (40%) or “usual material”Ever had sex 84.0%12–14 months, median 10.9Adherence: 70.8%33.2% (249/751)6 out of 7Zulaika 202333Kenya2017–2021Cluster-randomised trial4137 girls (median: 17.1, IQR 16.3–18.0)MooncupSexually active: 28.2%Non-cup user: Pads (61%) or “usual material”Sexually active: 27.1%2–3 years, median 2.5Adherence: 79.3%12.4% (514/4137)5 out of 7Mehta 202334,bKenya2018–2021Cluster-randomised trial436 (median: 16.9)MooncupSexually active: 35.6%Non-cup user: Pads 72%, cloth 3%, pads and cloth 25%Sexually active: 31.4%30 monthsAdherence estimated at 79.3%3.2% (14/436)6 out of 7Tessandier 202335France2016–2020Retrospective cohort138 women, 18–25 years (intervention median: 23.0; control: 21.0)Menstrual cup 75% of the timeIntercourse last 2 weeks: 72.9%Tampon 75% of the time. Intercourse last 2 weeks: 71.0%24 months (assessment of menstrual item use)Unclear6 out of 11Sica 202236USANRIndividually randomised cross-over trial65 women, 18–55 years (mean 34.2, SD NA, median 33, range 19–49)Tampax cup and another cupBaselineTampon 92%Cup 8%Baseline and 2 cycles (1 cycle with each product)f15.4% (10/65)6 out of 7Graciolli 202337Brazil2019Prospective cohort47 participants, 18 yrs and older (mean: 24.0, SD 4.7).Sexually active: 91.5%MooncupDisposable pads3 cycles. Adherence 97.9%Unclear6 out of 11Femme International38,cTanzania2019–2020Prospective cohort79 women, 20–50 yrs, mean 33, SD NALunette cupsNon-cup user (materials not specified)Cup use 2–3 yearsfNot reported4 out of 11Haahr 202539Denmark2017–2022Survey in 4 IVF-clinics1153 IVF patients, 18–42 years (mean 32.0, SD 5)No brand reportedNon-cup user:Pads (46%), tampons (29%), pads & tampons (24%)NAf1153/1511 available (76.3%)d6 out of 11Lebeer 202321Belgium2020–2020Cross-sectional study3043 women, ≥18 yrs, median ≈29 yrsNo brand reportedNon-cup user:Pads, tampons, liners & combinations.Percentage NANAf4.5% (149/3345)e6 out of 11Cohort studies or cross-sectional studies with insufficient quantitative information but statements on infectionsKarnaky 196240USANRProspective cohortsCohort 1: 50Cohort 2: 97Age not reportedTassette (vaginal)Non cup user: pads, tampons. Percentage NA24 months?UnclearfUnclear1 out of 11IQR, interquartile range; IVF, in-vitro fertilization; NA, not available; Pads, sanitary towels; SD, standard deviation; Yrs, years.Adherence: verbal report of menstrual cup use. Not available for Mehta and colleagues (2023), information used from Zulaika 2023.aA tool from the Cochrane Collaboration was used for trials (maximum of 7 points) and from the Joanna Briggs Institute for observational studies (maximum of 11 points).bMehta and colleagues, 2023, was a sub-study of Zulaika and colleagues, 2023, including 6 schools and only two study arms (menstrual cups vs. control) and more intense follow up (6, 12, 18 and 30 months) without overlap in infections reported and was considered a separate study.cData from this study were extracted as our best estimates from a figure in the publication. A request for more information about this study did not result in the acquisition of better data.dThe question about menstrual product use was not started from the beginning of the study.e4682 women registered with questionnaire, and 3345 women sent a self-swab vaginal sample. However, 149 out of 3345 samples did not pass the quality test.fNo information on sexual activity of participants or about menstrual cup adherence in this study. Studies contributing to this review. IQR, interquartile range; IVF, in-vitro fertilization; NA, not available; Pads, sanitary towels; SD, standard deviation; Yrs, years. Adherence: verbal report of menstrual cup use. Not available for Mehta and colleagues (2023), information used from Zulaika 2023. A tool from the Cochrane Collaboration was used for trials (maximum of 7 points) and from the Joanna Briggs Institute for observational studies (maximum of 11 points). Mehta and colleagues, 2023, was a sub-study of Zulaika and colleagues, 2023, including 6 schools and only two study arms (menstrual cups vs. control) and more intense follow up (6, 12, 18 and 30 months) without overlap in infections reported and was considered a separate study. Data from this study were extracted as our best estimates from a figure in the publication. A request for more information about this study did not result in the acquisition of better data. The question about menstrual product use was not started from the beginning of the study. 4682 women registered with questionnaire, and 3345 women sent a self-swab vaginal sample. However, 149 out of 3345 samples did not pass the quality test. No information on sexual activity of participants or about menstrual cup adherence in this study. Menstrual cup use and laboratory-confirmed infections and vaginal community state Results from a randomised controlled trial and pilot study among Kenyan schoolgirls were pooled; these studies compared STIs and bacterial vaginosis among menstrual cup users and non-cup users (control group, Fig. 2). A lower prevalence of STIs was seen in the adjusted estimates and overall, cup use was associated with a reduced risk of STIs (pooled adjusted prevalence ratio [aPR] 0.74, 95% CI 0.60–0.91, p = 0.0037 z-test, n = 695, I2 = 0.0%, Fig. 2). No significant difference by cup use was seen for N. gonorrhoeae and T. vaginalis, whereas a protective effect was seen for the adjusted estimate of C. trachomatis. The pooled aPR between cup use and bacterial vaginosis was 0.81 (95% CI 0.63–1.04, p = 0.0963 z-test, n = 693, I2 = 0.0%, Fig. 2) using data from two trials in Kenya.Fig. 2Pooled prevalence ratio for infections in randomised studies of menstrual cups, Kenya 2012–2021. CI, confidence interval; I2, I-squared, indicator of heterogeneity; NA, not available; Phillips-H, Phillips–Howard. ∗The model used by Mehta and colleagues was adjusted for repeated measurements, for the four measurement points for bacterial vaginosis and the two time points for sexually transmitted infections, with sample sizes varying by time of measurement.34 The sample size indicated is the minimum sample size at a time point. At baseline, 43 out of 436 participants (9.9%) had one or more STIs and 49 out of 436 participants (11.2%) had bacterial vaginosis. Baseline STI information was not available for Phillips–Howard and colleagues (2016). Notes: Each line indicates a unique study, and each outcome represents a meta-analysis model across the various indicated studies with the “Overall” line showing the pooled prevalence ratio for that outcome. The X-axis is on a log10 scale. Z-test for adjusted estimates: p = 0.0037 for sexually transmitted infections, p = 0.0120 for Chlamydia trachomatis, and p = 0.0963 for bacterial vaginosis. Random effects analysis when heterogeneity was present: adjusted pooled prevalence ratio for sexually transmitted infections: 0.69 (0.46–1.02, p = 0.064), and unadjusted pooled prevalence ratio for Trichomonas vaginalis 0.67 (0.34–1.31, p = 0.24). In the study by Phillips–Howard and colleagues, adjustment was for age, socioeconomic status, and reported sexual activity (had sexual intercourse) at the start of the study.13 In the study by Mehta and colleagues, adjustment was for age, sexual activity, socioeconomic status, school WASH (water, sanitation, and hygiene) score, and study time, and models accounted for repeated visits via random effect for participant.34 Pooled prevalence ratio for infections in randomised studies of menstrual cups, Kenya 2012–2021. CI, confidence interval; I2, I-squared, indicator of heterogeneity; NA, not available; Phillips-H, Phillips–Howard. ∗The model used by Mehta and colleagues was adjusted for repeated measurements, for the four measurement points for bacterial vaginosis and the two time points for sexually transmitted infections, with sample sizes varying by time of measurement.34 The sample size indicated is the minimum sample size at a time point. At baseline, 43 out of 436 participants (9.9%) had one or more STIs and 49 out of 436 participants (11.2%) had bacterial vaginosis. Baseline STI information was not available for Phillips–Howard and colleagues (2016). Notes: Each line indicates a unique study, and each outcome represents a meta-analysis model across the various indicated studies with the “Overall” line showing the pooled prevalence ratio for that outcome. The X-axis is on a log10 scale. Z-test for adjusted estimates: p = 0.0037 for sexually transmitted infections, p = 0.0120 for Chlamydia trachomatis, and p = 0.0963 for bacterial vaginosis. Random effects analysis when heterogeneity was present: adjusted pooled prevalence ratio for sexually transmitted infections: 0.69 (0.46–1.02, p = 0.064), and unadjusted pooled prevalence ratio for Trichomonas vaginalis 0.67 (0.34–1.31, p = 0.24). In the study by Phillips–Howard and colleagues, adjustment was for age, socioeconomic status, and reported sexual activity (had sexual intercourse) at the start of the study.13 In the study by Mehta and colleagues, adjustment was for age, sexual activity, socioeconomic status, school WASH (water, sanitation, and hygiene) score, and study time, and models accounted for repeated visits via random effect for participant.34 In Fig. 3, information from observational studies from the USA, Denmark and Tanzania were added to the trial results for bacterial vaginosis; the new unadjusted pooled estimate was 0.79 (95% CI 0.69–0.90, p = 0.0004 z-test, five studies, n = 2639, I2 = 0.0%, Fig. 3). A higher chance of a vaginal CST-I (dominated by L. crispatus) was noted among cup-users compared to non-users (pooled unadjusted PR 1.19, 1.11–1.27, four studies, p < 0.0001 z-test, n = 4255, I2 = 0.0%, Fig. 3). A higher adjusted mean relative abundance of L. crispatus was seen among girls using the menstrual cup (compared to non-cup users) in the only study which reported on this (adjusted mean relative abundance difference 4.46%, 95% CI 2.76–6.16%, Appendix p14).34 No clear effect of menstrual cup use was noted on C. albicans (three studies, Fig. 3).Fig. 3Associations between menstrual cup use and bacterial vaginosis, Candida albicans, and vaginal community state type 1, trials and cohort studies. CI, confidence interval; cRCT, cluster-randomised controlled trial; cRCp, cluster-randomised controlled pilot study; CST-I, vaginal community state type I; this is the optimal vaginal community state and a prevalence ratio above 1 indicates that this state is more likely, and not that infections are more likely. Int, international; NA, not available; Phillips-H, Phillips–Howard. ∗The model used by Mehta and colleagues (2023) adjusted for repeated measurements, for the four measurement points for bacterial vaginosis and the two time points for STIs, with sample sizes varying by time of measurement.34 The sample size indicated is the minimum sample size at a time point.† In the study by Haahr and colleagues, bacterial vaginosis was defined as the presence of aerobic vaginitis microbiota, an outcome which closely resembled Nugent score-positive bacterial vaginosis and was associated with poorer pregnancy outcomes in previous studies of the same group.41Notes: Each line indicates a unique study, and each outcome represents a meta-analysis model across the various indicated studies with the “Overall” line showing the pooled prevalence ratio for that outcome. The X-axis is on a log10 scale. Z-test for unadjusted estimates: p = 0.0004 for bacterial vaginosis (p = 0.0083 for adjusted estimate, 3 studies, I2 = 0.0%), p < 0.0001 for CST-I, and p = 0.91 for Candida albicans. In the study by Phillips–Howard and colleagues, adjustment was for age, household socioeconomic status, and reported sexual debut at the start of the study.13 In the study by Mehta and colleagues, adjustment was for age, sexual activity, household socioeconomic status, and school water, sanitation, and hygiene score, and study time, and models accounted for repeated visits via random effect for participant.34 In the study by Haahr and colleagues, adjustment was for body mass index, relationship status, smoking, and alcohol as binary variables.39 In the study by Lebeer and colleagues, adjustment was for age, sexual intercourse in the past 24 h and antibiotics in the past three months.21 Associations between menstrual cup use and bacterial vaginosis, Candida albicans, and vaginal community state type 1, trials and cohort studies. CI, confidence interval; cRCT, cluster-randomised controlled trial; cRCp, cluster-randomised controlled pilot study; CST-I, vaginal community state type I; this is the optimal vaginal community state and a prevalence ratio above 1 indicates that this state is more likely, and not that infections are more likely. Int, international; NA, not available; Phillips-H, Phillips–Howard. ∗The model used by Mehta and colleagues (2023) adjusted for repeated measurements, for the four measurement points for bacterial vaginosis and the two time points for STIs, with sample sizes varying by time of measurement.34 The sample size indicated is the minimum sample size at a time point.† In the study by Haahr and colleagues, bacterial vaginosis was defined as the presence of aerobic vaginitis microbiota, an outcome which closely resembled Nugent score-positive bacterial vaginosis and was associated with poorer pregnancy outcomes in previous studies of the same group.41Notes: Each line indicates a unique study, and each outcome represents a meta-analysis model across the various indicated studies with the “Overall” line showing the pooled prevalence ratio for that outcome. The X-axis is on a log10 scale. Z-test for unadjusted estimates: p = 0.0004 for bacterial vaginosis (p = 0.0083 for adjusted estimate, 3 studies, I2 = 0.0%), p < 0.0001 for CST-I, and p = 0.91 for Candida albicans. In the study by Phillips–Howard and colleagues, adjustment was for age, household socioeconomic status, and reported sexual debut at the start of the study.13 In the study by Mehta and colleagues, adjustment was for age, sexual activity, household socioeconomic status, and school water, sanitation, and hygiene score, and study time, and models accounted for repeated visits via random effect for participant.34 In the study by Haahr and colleagues, adjustment was for body mass index, relationship status, smoking, and alcohol as binary variables.39 In the study by Lebeer and colleagues, adjustment was for age, sexual intercourse in the past 24 h and antibiotics in the past three months.21 In Fig. 4, time patterns for STIs can be seen. In the study by Phillips–Howard and colleagues, effect modification was noted by time for STIs13; among participants that used the cup for nine or more months, fewer STIs were noted in the cup group compared to controls but not before (Fig. 4). C. trachomatis was significantly lower among cup-users at 30 months, but not at 12 months in the study by Mehta and colleagues.34 A trial in Kenya reported on HIV and HSV-2 incidence following cup use compared to that in participants in the control arm.33 The number of new HIV infections were low, and slightly lower among cup users than non-cup users (1.0% vs. 1.3% for per protocol population, p > 0.05, Fig. 4). The same pattern was seen for HSV-2 (8.8% among cup users vs. 12.1% among non-cup users for per protocol population), with an aPR of 0.71 (95% CI 0.50–1.02, intention-to-treat analysis). No difference was seen for human papilloma virus by menstrual cup use in one study by Tessandier and colleagues (Fig. 4).35Fig. 4Associations between menstrual cup use and outcomes related to sexually transmitted infections: time trends and additional information available. CI, confidence interval; cRCT, cluster-randomised controlled trial; HIV, human immunodeficiency virus; HSV-2, herpes simplex virus type 2; ITT, intention-to-treat analysis; PP, per protocol analysis; NA, not available; Phillips-H, Phillips–Howard. Notes: Each line indicates a time point, or a different type of analysis for a study. In this forest plot there are no pooled estimates. The X-axis is on a log10 scale. In the study by Phillips–Howard and colleagues, adjustment was for age, household socioeconomic status, and reported sexual debut at the start of the study.13 In the studies by Zulaika and colleagues and Mehta and colleagues, adjustment was for age, site, school water, sanitation, and hygiene score, baseline HIV and HSV-2 status, household socioeconomic status, and reported sexual debut at baseline.33,34 In the study by Mehta and colleagues, there was additionally adjustment for study time, and models accounted for repeated visits via random effect for participant. In the study by Zulaika and colleagues, at baseline 1.2% (13 girls) and 2.0% (19 girls) were infected with HIV and 16.5% (171 girls) and 18.9% (174 girls) were infected with HSV-2 in the menstrual cup and control arm, respectively. Associations between menstrual cup use and outcomes related to sexually transmitted infections: time trends and additional information available. CI, confidence interval; cRCT, cluster-randomised controlled trial; HIV, human immunodeficiency virus; HSV-2, herpes simplex virus type 2; ITT, intention-to-treat analysis; PP, per protocol analysis; NA, not available; Phillips-H, Phillips–Howard. Notes: Each line indicates a time point, or a different type of analysis for a study. In this forest plot there are no pooled estimates. The X-axis is on a log10 scale. In the study by Phillips–Howard and colleagues, adjustment was for age, household socioeconomic status, and reported sexual debut at the start of the study.13 In the studies by Zulaika and colleagues and Mehta and colleagues, adjustment was for age, site, school water, sanitation, and hygiene score, baseline HIV and HSV-2 status, household socioeconomic status, and reported sexual debut at baseline.33,34 In the study by Mehta and colleagues, there was additionally adjustment for study time, and models accounted for repeated visits via random effect for participant. In the study by Zulaika and colleagues, at baseline 1.2% (13 girls) and 2.0% (19 girls) were infected with HIV and 16.5% (171 girls) and 18.9% (174 girls) were infected with HSV-2 in the menstrual cup and control arm, respectively. Remaining infections and time trends that could not be pooled are reported in Appendix p15. For bacterial vaginosis, a protective trend was generally seen with longer duration of cup use, whereas the prevalence ratio for CST-I remained in the same direction at different time points. One study reported no statistical difference in the prevalence of S. aureus between cup (9.6%) and control (11.3%) arms (PR 0.85, 95% CI 0.46–1.56, Appendix p15) among 363 participants.13 In the same study (n = 188 cup users) and in the study by Zulaika and colleagues (n = 2119 cup users), no cases of toxic shock syndrome were reported among menstrual cup users.13,33 Unquantifiable or additional information on infections among menstrual cup users are available in Appendix p16. No increased risk of any of the infections evaluated were observed. A large community-based survey in Belgium by Lebeer and colleagues (2023, n = 3043 participants) noted that the menstrual cup was associated with the L. crispatus module (a group of correlated bacterial subgenera with L. crispatus as the most abundant genus; aOR 1.17, 95% CI 1.08–1.27, p = 0.0035).21 In this study, CST-I was split in I-A and I-B with communities with mostly L. crispatus in CST I-A and communities with lower, moderate relative abundance of L. crispatus in I–B; the adjusted ORs for menstrual cup use and these community states were 1.14 (95% CI 0.95–1.37, p = 0.62) and 1.13 (95% CI 0.86–1.48, p = 0.82), respectively. Because of the small number of studies per infection, only one subgroup analysis by study quality was conducted for menstrual cup use and bacterial vaginosis (Appendix p17); pooled estimates of two studies of high (PR 0.81, 0.70–0.95, p = 0.0070, n = 693) and three studies of low quality (PR 0.71, 0.54–0.94, p = 0.0151, n = 1946) were similar (meta-regression p = 0.47). For bacterial vaginosis, a leave-one-out analysis showed similar results for the pooled estimate when removing studies one by one, and there was no indication of funnel plot asymmetry (Appendix p18) but number of studies were small. Results for infections unconfirmed by laboratory tests and a provisional table of evidence for laboratory confirmed infections using GRADE are available in the Appendix (p19 & 20, respectively). Discussion In this systematic review and meta-analysis, we assessed the effect of menstrual cups on STIs, RTIs and the vaginal microbiome to aggregate evidence to inform programmes and individuals wanting to use the menstrual cup. We identified eleven studies which reported on outcomes of interest. Our analyses demonstrated that menstrual cups were associated with biological benefits, including lower proportions of bacterial vaginosis, with effects seen in populations across continents and age groups. Among schoolgirls in Kenya, STIs and HSV-2 infection were lower among cup users. Additionally, a beneficial effect on maintaining a healthy vaginal microbiome was observed in different geographic regions. Menstrual cups were not associated with an increased risk in any of the infections evaluated. The importance of a healthy vaginal microbiome to protect from vaginal infections has become better established in recent years due to advances in molecular tools.42 The vaginal microbiome is typically categorised into five CSTs, and vaginal microbial composition can fluctuate for short periods of time,43,44 as well as experience durable shifts across different life stages.45 Numerous factors are known to affect the vaginal microbiome composition and stability, including genetics, endogenous hormones, contraception, sexual behaviours, and personal hygiene.44,46 Oestrogens stimulate vaginal epithelial cells to produce glycogen, which is a source of the cell-free glucose that lactobacilli ferment into lactic acid.46,47 Lactic acid provides an acidic vaginal environment that is not conducive for STIs.46 Lactobacilli also produce hydrogen peroxide (H2O2) and bacteriocins and compete with STI pathogens for nutrients and adherence to vaginal surface epithelium.46 The oxidising property of H2O2 can additionally damage invading pathogens.26 Menstruation can modify the vaginal microbiome.48 The lysis of vaginal blood during menses and increased iron levels can support the accelerated growth of other bacteria, such as Gardnerella vaginalis and L. iners.49 Menstrual cup usage may affect the vaginal microbiome by reducing contact of the microbiome with blood (because it is collected in the cup) which may promote faster vaginal microbiome recovery or maintenance of healthy vaginal microbiome during menses.34 An American study which measured vaginal pH during the cycle, observed no changes in vaginal pH during menstruation among menstrual cup users.40 Beneficial SRH effects were noted in this meta-analysis for both low- and high-income countries, across different comparators (e.g., reusable or disposable pads, tampons), and across different age groups. Therefore, observed SRH impacts cannot be solely due to factors like poor hygiene conditions or sub-optimal menstrual product comparators. L. crispatus offers protection against vaginal dysbiosis and STIs while disruption of the microbial balance increases susceptibility to a variety of STIs, including C. trachomatis, N. gonorrhoeae, T. vaginalis, and human papillomavirus.24 Bacterial vaginosis is the most common vaginal infection, and has been associated with adverse birth outcomes, infertility, increased risk of HIV/STI acquisition, post-operative gynaecological infections and pelvic inflammatory disease.50 In countries with high burdens of HIV, bacterial vaginosis has been consistently associated with HIV acquisition,51 and vaginal concentrations of pro-inflammatory cytokines and chemokines identify women at increased risk of HIV infection.52 The results of the five included studies on bacterial vaginosis from three continents all point to a protective effect of the menstrual cup, with estimates in the two randomised studies having similar strengths when adjusted for socio-economic status and sexual behaviour (Fig. 2), and generally over time the same protective direction (Appendix p15). C. trachomatis was the most consistently affected STI and is known to be susceptible to lactic acid in the vagina.53 However, the number of studies with information on STIs was low. This review has several limitations. Although eleven studies were identified for inclusion, one could not be included in the meta-analysis due to insufficient quantitative data. Similarly, due to a lack of uniformity in study outcomes, few studies could be pooled. Nonetheless, where pooled, heterogeneity was low (I2 <35%) indicating robustness of effect. Despite the low heterogeneity, there may have been selection bias in the study by Haahr and colleagues, as the authors indicated that more samples for microbiome characterisation were missing from women with normal vaginal microbiota.39 The quality of studies included in the meta-analysis was moderate, with only three identified as good quality, potentially contributing to bias in the analysis. However, results did not change when including low quality studies. We focused on studies with information on menstrual cup use and laboratory confirmed infections in this review and did not evaluate participants' experience of menstrual cup use, including the presentation of clinical symptoms. Of the eleven studies included, five used self-collected vaginal swabs as the sampling method. While a practical and financially feasible sampling method, this could result in lower sensitivity if the samples were not collected properly. However, several studies have shown that self-collected swabs had comparable diagnostic accuracy to healthcare worker-collected samples for STIs, bacterial vaginosis, candidiasis, and vaginal microbiome54,55; e.g. in a review by Jaya et al. (2024) concordance (Cohen's kappa) between self-collected swab and healthcare worker collected swab for C. trachomatis and T. vaginalis was >90% (3 and 4 studies, respectively), for N. gonorrhoeae this was >85% (3 studies).54 Vaginal sampling approach has been shown to have minimal contribution to microbial community structure analysis,56,57 nor do a variety of conditions for timing of processing (e.g., immediate vs. stored) or duration of storage.58 DNA extraction protocol can affect DNA yield, genomic quality score, microbial DNA abundance, alpha diversity, and observed microbial community structure.59 Choosing which portion of the microbial 16 S ribosomal RNA (rRNA) gene to target using PCR can impact the characterisation of the microbial community in samples. This is due to the presence of nine different ‘hypervariable’ regions of the gene. Each hypervariable region can provide different levels of taxonomic resolution for microbial community analysis. Historically, PCR primers amplifying the third and fourth hypervariable regions (i.e., V3/V4) have been most frequently employed in vaginal microbiome studies, as these have provided good species-level resolution for critical vaginal microbial taxa.60 Some studies have shown that primers targeting the V3/V4 region have shown better performance than primers targeting the V1/V2 region,60,61 while others have shown relative insensitivity to variable region chosen (e.g., France et al., 2020).62 Even though these sample processing stages can introduce variability in observed microbial communities, we observed little heterogeneity in our analyses, and consistent effects across studies, indicating the robustness of the association. However, the certainty of evidence for outcomes was assessed as low and very low, indicating that more evidence is needed to strengthen the results (Appendix p20). To note, the observed effects of menstrual cup use on STIs come from studies among schoolgirls in western Kenya, where menstrual product use and infection risk may differ from other settings, potentially limiting global generalisability. Family planning methods, including use of hormonal contraceptives, could be a potential confounder of the association between menstrual cup and vaginal microbiome; however, there was insufficient information to take this into account. Whether people who menstruate with more optimal vaginal microbiome were disproportionately represented in studies reporting on healthy vaginal microbiome and reduced infection prevalence following menstrual cup use is unclear. Lastly, all studies relied on a verbal report of adherence to cup use, with one study exploring cup colour change to evaluate use.63 This systematic review found that menstrual cup use conferred protective effects on STIs in Kenya and was associated with beneficial impacts on the vaginal microbiome and bacterial vaginosis across different regions, potentially supporting overall SRH. Many studies have reported on the association between bacterial vaginosis and preterm birth or HIV. A recent review by Hadhoum et al. (2025) reported a pooled odds ratio of 1.60 (95% CI 1.36–1.89, I2 67%, 28 studies, 50,566 participants) for the association between bacterial vaginosis and preterm birth.23 A review by Hilber et al. (2010) reported a pooled adjusted rate ratio of 1.57 (95% CI 1.26–1.94, I2 19%, 8 studies) for the association between bacterial vaginosis and HIV-infection among longitudinal studies.64 Promoting healthy vaginal microbiomes through safe and effective menstrual hygiene practices with use of menstrual cups offers a low-cost, scalable and sustainable solution to reducing the burden of preventable gynaecological conditions.65 These benefits are particularly significant in low-resource settings, where limited access to healthcare heightens the risk of adverse outcomes associated with vaginal dysbiosis and infection. Menstrual cups are cost-effective, safe and durable, and accepted among diverse populations.17,66 The upfront costs of a menstrual cup compared to alternative menstrual products and the limited availability in low-income and middle-income countries may constrain equity. The inclusion of menstrual cups in national menstrual health policies, including provision through schools, would facilitate use and equity of the menstrual cup.17 Our findings strengthen evidence for policymakers and health programmes that menstrual cups are a multipurpose tool, supporting menstrual care and addressing period poverty while also showing effects on the vaginal microbiome and SRH. The menstrual cup is an effective and environmentally friendly tool to manage menstruation, and relative to other methods of menstrual hygiene management, can result in reduced risks of STIs, and can lower the prevalence of bacterial vaginosis and support a higher prevalence of optimal vaginal microbiome composition. Further research is needed to strengthen the biological evidence across socioeconomic and geographic contexts and assess the mechanism of action on SRH. Contributors PAPH and SDM conceptualised the study. AMvE and AQ-S wrote the protocol, and all authors reviewed the protocol. AMvE and AQ-S screened records and extracted data, assessed the quality of included studies and verified the data. AMvE and GZ conducted the statistical analyses. AMvE and AQ-S wrote the first draft of the manuscript, and all authors (GZ, AS, LM, SJG, PAPH and SDM) read, provided critical input, interpreted the data analyses, and critically revised and approved the final version of the manuscript. All authors had full access to all the data in the study. The corresponding author (AMvE) together with the joint senior authors (PAPH, SDM) had final responsibility for the decision to submit for publication. Data sharing statement This study is based on published data from primary studies, supplemented with information from authors of primary studies where needed. Data is available in the articles, on the web (Zulaika et al., 2023), at the discretion of the authors (Phillips–Howard et al., 2016; Mehta et al., 2023; Haahr et al., 2025) or in a repository (Lebeer et al., 2023). The dataset with extracted information and the Stata code for all figures are available in figshare (doi 10.6084/m9.figshare.32145175 and doi 10.6084/m9.figshare.32145136, respectively). Declaration of interests We declare no competing interests.