chapter 3 research methodology statistical treatment of data

Academia.edu no longer supports Internet Explorer.

To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to  upgrade your browser .

Enter the email address you signed up with and we'll email you a reset link.

• We're Hiring!
• Help Center

Chapter 3 RESEARCH AND METHODOLOGY

Related Papers

Jessa Arcina

Research design is the blue print of the procedures that enable the research to test hypothesis by reaching valid conclusions about the relationship between dependent and in depend variables. It is a plan structure and strategy of research prepared to obtain answer to research questions and to control variances. Before doing the various studies on the present thesis the researcher has fixed the topic and area because it provide the entire draft of the scheme of the research staring from writing the hypothesis their operational implications to the final analysis of the data. The structural of the research is more specific as it provides the outline, the scheme the paradigm o f the operation of the variables. It presents a series of guide posts to enable the researcher to progress in the right direction it gives an outline of the

Scholarly Communication and the Publish or Perish Pressures of Academia A volume in the Advances in Knowledge Acquisition, Transfer, and Management (AKATM) Book Series

Dr. Naresh A . Babariya , Alka V. Gohel

The most important of research methodology in research study it is necessary for a researcher to design a methodology for the problem chosen and systematically solves the problem. Formulation of the research problem is to decide on a broad subject area on which has thorough knowledge and second important responsibility in research is to compare findings, it is literature review plays an extremely important role. The literature review is part of the research process and makes a valuable contribution to almost every operational step. A good research design provides information concerning with the selection of the sample population treatments and controls to be imposed and research work cannot be undertaken without sampling. Collecting the data and create data structure as organizing the data, analyzing the data help of different statistical method, summarizing the analysis, and using these results for making judgments, decisions and predictions. Keywords: Research Problem, Economical Plan, Developing Ideas, Research Strategy, Sampling Design, Theoretical Procedures, Experimental Studies, Numerical Schemes, Statistical Techniques.

Xochitl Ortiz

The authors felt during their several years of teaching experience that students fail to understand the books written on Research Methodology because generally they are written in technical language. Since this course is not taught before the Master’s degree, the students are not familiar with its vocabulary, methodology and course contents. The authors have made an attempt to write it in very non- technical language. It has been attempted that students who try to understand the research methodology through self-learning may also find it easy. The chapters are written with that approach. Even those students who intend to attain high level of knowledge of the research methodology in social sciences will find this book very helpful in understanding the basic concepts before they read any book on research methodology. This book is useful those students who offer the Research Methodology at Post Graduation and M.Phil. Level. This book is also very useful for Ph.D. Course Work examinations.

Chisomo Mgunda

MD Ashikur Rahman

Roxannie Ibot

Second Language Learning and Teaching

Magdalena Walenta

Lester Millara

the purpose of this paper is to know what are the use of methodology in a research paper.

Asmatullah Ghayasi

Ralph Renzo Salangsang

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

RELATED PAPERS

collins wetiatia

Dr Sunarsih

Naeem Tabassum

Dhea Savella Segui

melaku emayneh

Godwill Medukenya

Ponsian P R O T Ntui

Satyajit Behera

Dr. Moses Gweyi

Miyoko Chin

Sasmitha Pramudya Machmud Machmud

Binu Peniel

Sadman Sakib

Anil Jharotia

Yuanita Damayanti

Vikalpa: The Journal for Decision Makers

Vivek Patkar

Plastic and reconstructive surgery. Global open

Rolf Gemperli

Khamis S Moh'd

Williamson Sitorus

Journal of Xidian University

Rahat Sabah

Bakhtawer Zain

•   We're Hiring!
•   Help Center
• Find new research papers in:
• Health Sciences
• Earth Sciences
• Cognitive Science
• Mathematics
• Computer Science
• Academia ©2024

Methods Section: Chapter Three

The methods section , or chapter three, of the dissertation or thesis is often the most challenging for graduate students.  The methodology section, chapter three should reiterate the research questions and hypotheses, present the research design, discuss the participants, the instruments to be used, the procedure, the data analysis plan , and the sample size justification.

Research Questions and Null Hypotheses

Chapter three should begin with a portion that discusses the research questions and null hypotheses.  In the research questions and null hypotheses portion of the methodology chapter, the research questions should be restated in statistical language.  For example, “Is there a difference in GPA by gender?” is a t-test type of question, whereas “Is there a relationship between GPA and income level?” is a correlation type of question.  The important thing to remember is to use the language that foreshadows the data analysis plan .  The null hypotheses are just the research questions stated in the null; for example, “There is no difference in GPA by gender,” or “There is no relationship between GPA and income level.”

Research Design

The next portion of the methods section, chapter three is focused on developing the research design.  The research design has several possibilities. First, you must decide if you are doing quantitative, qualitative, or mixed methods research. In a quantitative study, you are assessing participants’ responses on a measure.  For example, participants can endorse their level of agreement on some scale.  A qualitative design is a typically a semi-structured interview which gets transcribed, and the themes among the participants are derived.  A mixed methods project is a mixture of both a quantitative and qualitative study.

Participants

In the research methodology, the participants are typically a sample of the population you want to study.  You are probably not going to study all school children, but you may sample from the population of school children.  You need to include information about the characteristics of the population in your study (Are you sampling all males? teachers with under five years of experience?).  This represents the participants portion of your methods section, chapter three.

Need help with your methods section?

Schedule a time to speak with an expert using the calendar below.

Instruments

The instruments section is a critical part of the methodology section, chapter three.  The instruments section should include the name of the instruments, the scales or subscales, how the scales are computed, and the reliability and validity of the scales.  The instruments portion should have references to the researchers who created the instruments.

The procedure section of the methods chapter is simply how you are going to administer the instruments that you just described to the participants you are going to select.  You should walk the reader through the procedure in detail so that they can replicate your steps and your study.

Data Analysis Plan

The data analysis plan is just that — how you are going to analyze the data when you get the data from your participants.   It includes the statistical tests you are going to use, the statistical assumptions of these tests, and the justification for the statistical tests.

Sample Size Justification

Another important portion of your methods chapter three, is the sample size justification.  Sample size justification (or power analysis) is selecting how many participants you need to have in your study.  The sample size is based on several criteria:  the power you select (which is typically .80), the alpha level selected (which is typically .05), and the effect size (typically, a large or medium effect size is selected).  Importantly, once these criteria are selected, the sample size is going to be based on the type of statistic: an ANOVA is going to have a different sample size calculation than a multiple regression.

Statistical Methods

Cite this chapter.

• Paul Taylor 3  

584 Accesses

3 Citations

3.5. Conclusion

The techniques presented in this chapter do not form anything like an exhaustive list of useful statistical methods. These techniques were chosen because they are either widely used or ought to be widely used. The regression techniques are widely used, though there is some reluctance amongst researchers to make the jump from linear models to generalized linear models.

The multivariate analysis techniques ought to be used more than they are. One of the main obstacles to the adoption of these techniques may be that their roots are in linear algebra.

I feel the techniques presented in this chapter, and their extensions, will remain or become the most widely used statistical techniques. This is why they were chosen for this chapter.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

• Get 10 units per month
• Download Article/Chapter or Ebook
• 1 Unit = 1 Article or 1 Chapter
• Cancel anytime
• Available as PDF
• Read on any device
• Instant download
• Own it forever
• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info
• Durable hardcover edition

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Unable to display preview.  Download preview PDF.

Similar content being viewed by others

General Aspects of Fitting Regression Models

Introduction and Preview

Multivariate Exploratory Approaches

Author information, authors and affiliations.

University of Hertfordshire, UK

Paul Taylor

You can also search for this author in PubMed   Google Scholar

Editor information

Editors and affiliations.

FB Informatik und Informationswissenschaft, Universität Konstanz, 78457, Konstanz, Germany

Michael Berthold

Department of Mathematics, Imperial College, Huxley Building 180 Queen’s Gate, London, SW7 2BZ, UK

David J. Hand

Rights and permissions

Reprints and permissions

Copyright information

© 2007 Springer-Verlag Berlin Heidelberg

About this chapter

Taylor, P. (2007). Statistical Methods. In: Berthold, M., Hand, D.J. (eds) Intelligent Data Analysis. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-48625-1_3

Download citation

DOI : https://doi.org/10.1007/978-3-540-48625-1_3

Publisher Name : Springer, Berlin, Heidelberg

Print ISBN : 978-3-540-43060-5

Online ISBN : 978-3-540-48625-1

eBook Packages : Springer Book Archive

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

• Open access
• Published: 18 June 2024

Strong evidence supports the use of estradiol therapy for the treatment of vaginal inflammation: a two-way Mendelian randomization study

• Xiaosheng Xu 1 ,
• Yan Liu 1 ,
• Weiwei Feng 1 &
• Jian Shen 1  

European Journal of Medical Research volume  29 , Article number:  339 ( 2024 ) Cite this article

250 Accesses

Metrics details

Nowadays, there has been limited Mendelian randomization (MR) research focusing on the causal relationship between estradiol and vaginitis. Therefore, this study conducted a two-way MR study to clarify the causal effect and related influencing factors between them.

All genetic datasets were obtained using publicly available summary statistics based on individuals of European ancestry from the IEU GWAS database. MR analysis was performed using MR-Egger, weighted median (WM) and inverse variance weighted (IVW) methods to assess the causal relationship between exposure and outcome and to validate the findings by comprehensively evaluating the effects of pleiotropic effects and outliers.

MR analysis revealed no significant causal relationship between estradiol and vaginitis risk. There was a negative correlation between estradiol and age at menarche (IVW, OR: 0.9996, 95% CI: 0.9992–1.0000, P  = 0.0295; WM, OR: 0.9995, 95% CI: 0.9993–0.9998, P  = 0.0003), and there was a positive correlation between age at menarche and vaginitis (IVW, OR: 1.5108, 95% CI: 1.1474–2.0930, P  = 0.0043; MR-Egger, OR: 2.5575, 95% CI: 1.7664–9.6580, P  = 0.0013). Estradiol was negatively correlated with age at menopause (IVW, OR: 0.9872, 95% CI: 0.9786–0.9959, P  = 0.0041). However, there was no causal relationship between age at menopause and vaginitis ( P  > 0.05). In addition, HPV E7 Type 16, HPV E7 Type 18, and Lactobacillus had no direct causal effects on estradiol and vaginitis ( P  > 0.05). Sensitivity analyses revealed no heterogeneity and horizontal pleiotropy.

When estrogen levels drop, it will lead to a later age of menarche, and a later age of menarche may increase the risk of vaginitis, highlighting that the longer the female reproductive tract receives estrogen stimulation, the stronger the defense ability is formed, and the prevalence of vaginitis is reduced. In conclusion, this study indirectly supports an association between reduced level of estrogen or short time of estrogen stimulation and increased risk of vaginitis.

Introduction

Vaginitis is caused by infection, inflammation and an imbalance of normal flora. Typical symptoms include odor, itching, irritation, burning, dysuria, or changes in vaginal discharge [ 1 , 2 , 3 ]. Vaginitis can negatively impact women's quality of life, with some women reporting anxiety, shame, and hygiene concerns, especially those with recurrent symptoms [ 4 ]. Hypoestrogen (atrophic) individuals may present with symptoms of vaginitis and benefit from hormone therapy or replacement therapy [ 5 ]. At present, local estrogen therapy is mainly used in the clinical treatment of atrophic vaginitis (AV) [ 6 ]. A meta-analysis of 10 clinical trials showed that vaginal estrogen therapy resulted in the greatest relief of symptoms of vaginal dryness and improvement of vaginal/vulvar atrophy [ 7 ]. Estradiol, the main estrogen produced by the ovaries, plays a vital role in keeping vaginal tissues lubricated and healthy [ 8 ]. However, existing studies are heterogeneous due to the diversity of detection methods and sample types. There is no consistent conclusion yet on the factors through which estradiol may affect vaginitis. Therefore, new comprehensive analyses are urgently needed to improve our understanding of the relationship between estradiol and vaginitis.

In contrast to the gut, which is colonized by a complex microbiome, the human vaginal ecosystem is associated with low microbial diversity, mainly dominated by a small number of Lactobacillus species [ 9 , 10 ]. Microbial imbalances can manifest clinically as abnormal vaginal discharge with a "fishy" odor [ 11 ]. Human papovavirus (HPV) infection is common and spread by direct contact. Despite the high prevalence of both cervicovaginal HPV infection and vaginosis worldwide, their causal relationship remains unclear [ 12 ]. Menarche marks the beginning of the maturation and function of the female reproductive system, and approximately 50–80% of the age at puberty is determined by genetic factors [ 13 ]. Some studies have shown that early menarche may be related to factors such as obesity, high energy intake, and family environment, while late menarche may be related to factors such as emaciation and stress, among which estrogen also plays a key role [ 1 , 14 ]. As age increases, a significant decrease in estrogen can reduce vaginal lubrication and cause vaginal atrophy. Vaginal atrophy is the most common problem of menopause, with serious social, psychological and medical consequences. According to the North American Menopause Society, 40 percent of postmenopausal women may develop AV [ 15 ].

A randomized controlled trial (RCT) should be an ideal study design to confirm the causal relationship between estradiol and vaginitis. However, conducting randomized controlled trials in reality faces difficulties. MR minimizes the effects of measurement error and directional causality. Because these instrumental variables (IVs) remain constant after conception and are not expected to be affected by potential founder effects, the MR approach overcomes some of the limitations of traditional epidemiological studies. Mendelian randomization (MR), a method that uses genetic variation to measure causal exposure relationships among disease risk variables, can remove confounding bias inherent in observational studies [ 16 ]. Clinical studies have revealed that vaginitis is affected by many factors. However, so far, few MR studies have focused on their causal relationship. To elucidate the causal direction between estradiol, vaginitis, age at menarche, age at menopause, HPV and Lactobacillus , a two-way MR study was performed using the genome-wide association study (GWAS) database in this study.

Materials and methods

Study design.

To investigate the role of estradiol in vaginitis, we conducted Mendelian randomization (MR) analyses on two samples using GWAS data. Standard MR analysis requires that the following three model assumptions must be met [ 17 ]: (1) SNPs used as instrumental variables (IVs) are significantly associated with estradiol, vaginitis, age at menarche, age at menopause, HPV and Lactobacillus and reach the genome-wide significance threshold; (2) SNPs were independent of confounding factors; (3) SNPs were only associated with vaginitis through estradiol, age at menarche, age at menopause, Lactobacillus and HPV, but not through other pathways (Fig.  1 ). The first hypothesis can be tested directly using observational data. However, the last two assumptions are often difficult to test in practice. In this study, we validated our findings using the MR approach under different model assumptions.

Schematic showing the steps of the three-step dual-sample MR analysis

Data sources

The analysis in this study was based on the publicly available online database of the IEU Genome-Wide Association Studies (GWAS) ( https://gwas.mrcieu.ac.uk/ ) [ 18 ]. Details can be found in Table  1 . In addition, we provide a detailed SNP list and summary statistics from the original GWAS (Table S1).

Selection of instrumental variables

GWAS have fueled the rise of MR by identifying genetic variants that can be used as instrumental variables in a two-step framework to determine whether pending DNA methylation is present along a causal pathway between exposure and disease. In our two-sample MR, instrumental variables (IVs) were selected through the following rigorous procedure: (1) we selected SNPs that were significantly associated with exposure ( P  < 5 × 10 –8 , at the genome-wide threshold); (2) we obtained independent SNPs were identified, and the clustering threshold ( r 2  < 0.001, clump distance > 10,000 kb) in the PLINK clustering method was used to eliminate SNPs that caused deviations due to linkage disequilibrium (LD), and according to the allele frequency and incompatible alleles of the palindrome. Among them, low-quality SNPs were eliminated. To assess weak instrument bias, we calculated F statistics, and only SNPs with F  < 10 were considered weak and removed [ 19 ].

MR analysis

This study mainly used MR-Egger, weighted median (WM) and IVW (inverse variance weighted), a total of 3 algorithms for two-sample MR analysis to evaluate the causal relationship between estradiol and vaginitis. The MR-Egger approach provides a progressively consistent measure of causal effect, adjusting for horizontal pleiotropy by pooling individual SNP-specific Wald ratios via adaptive Egger regression. The WM method produces progressively consistent causal effect estimates by using the weighted median of the Wald ratios, provided that at least 50% of the variants meet the effective IV of the exclusion limits. IVW was identified as the primary method, the weighted linear regression model, in the absence of IV In the case of horizontal pleiotropy, the results have high confidence [ 19 ].

Sensitivity analysis

The results of the primary analysis were validated by several different MR approaches with different model assumptions: (1) level multiplicity occurs when the genetic variation associated with the exposure of interest directly affects the outcome through multiple pathways other than the assumed exposure. Effectiveness: MR Egger's intercept test was used to determine the level of pleiotropy among these SNPs [ 20 ]; (2) Cochran's Q statistic of the IVW method was used to assess the heterogeneity of IV causality estimates among individuals, if the Cochran's Q test's P Values < 0.05, heterogeneity was detected. The smaller heterogeneity indicates that MR estimation is more reliable; (3) leave-one-out (LOO) analysis can be used to evaluate the influence of single SNP causal estimation, and each exposure-related SNP is sequentially discarded by LOO analysis to repeat IVW analysis. In addition, we also drew diagnostic maps to illustrate the MR results. Funnel plots can exploit the symmetry of the graphical representation to visually detect directional pleiotropy and assess the robustness of the results.

Statistical analysis

MR is based on the principle of random distribution of genetic genes. When the frequency of SNPs is highly consistent with the changes in exposure variables, it can be preliminarily considered that SNPs are related to exposure variables. All statistical tests were two-sided and were considered to show statistical significance at a p -value < 0.05. The above analyses were performed using the "TwoSampleMR" package (Version: 0.5.6) and "MRPRESSO" in R software (Version: 4.1.0) " package (Version: 1.0). We first tested for heterogeneity between variant-specific estimates using Cochran Q statistic in the inverse-variance weighted model. We then explored horizontal pleiotropy via the MR-Egger and Outlier methods.

Selected SNPs and IVs validation

The identification and information of genetic variants associated with “Estradiol”, “HPV E7 Type 16”、 “HPV E7 Type18”、 “ Lactobacillus ”、 “Age when periods started (menarche)”, “Age at menopause”, and “Vaginitis” in this study were obtained from the GWAS database. In view of the GWAS data that are significantly associated with the above disease phenotypes, after removing the biased and low-quality SNPs caused by LD, 205、27、13、54、196、30, and 108 SNPs with each F statistic above 10 were retained as IVs ( P  < 5 × 10 –8 , r 2  < 0.001). Therefore, we could conclude that weak instrument bias would not substantially influence causal inference in our MR analysis.

Causal relationship between HPV、 Lactobacillus and estradiol and vaginitis

To assess the causal effect between estradiol and vaginal disease, MR analysis was used to infer a bidirectional causal relationship between the two. The results showed that there was a negative correlation trend between estradiol and vaginitis. However, the difference was not significant (Figures S1 and S2). Two-way MR analysis was performed on HPV E7 Type 16, HPV E7 Type 18, and genus Lactobacillus id.1837, respectively, with estradiol and vaginitis. Our MR results based on genetic data did not support a significant causal relationship between the above variables (Table S2).

Causal relationship between estradiol and menarche

Estrogen plays key role in early and late menarche age [ 1 , 14 ]. Therefore, we used the GWAS data of age when periods started (menarche) for analysis. The results showed that the OR [95% CI] obtained by the IVW algorithm was 1.0000 [0.9992–1.0000], the β value was -0.0004, and the test P value was 0.0295. However, the Cochran Q test in the IVW method showed that there was some heterogeneity among the selected SNPs ( Q _pval = 0.0283), which may be related to the small number of SNPs. The OR estimates between estradiol and age when periods started (menarche) were 1.0000 (95%CI 0.9993–0.9998; p  = 0.0003) for the weighted median method. The MR‐Egger intercepts showed no evidence for significant directional pleiotropy ( P intercept = 0.2283 for estradiol), suggesting that there is no directional pleiotropy in our two-sample MR. The above results show that the level of estradiol is negatively correlated with the age of menarche, and when estrogen decreases, it may lead to a delay in the age of menarche (Table  2 , Fig.  2 ). In two-way analyses, the menarche-estradiol causality model was also significant, suggesting reverse causality, since the age at menarche is a continuous variable, in the MR analysis, the OR should be based on the 1SD increase of the exposure variable, so we also provide the 1SD value of the age at menarche ( Figure S3 ). One standard deviation (SD) increases in age of menarche, 181,642.3 (95% CI 9.88825e−07 – 3.336679e + 16, P  = 0.3601256) for estradiol.

2-sample MR analysis. A Scatterplot of the causal effect of estradiol on age at menarche. The slope of the line indicates the magnitude of the causal relationship. B Funnel plot. C Forest plot of MR analysis representing causal estimates of estradiol on age at menarche. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple‐instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. D LOO analyses. The black dots represent one IVW and the red dots represent estimates using all IVs. Horizontal lines indicate 95% confidence intervals

Causal relationship between menarche and vaginitis

To assess the causal effect between menarche and vaginal disease, MR analysis was also used to infer a bidirectional causal relationship between the two. As shown in Table  3 and Fig.  3 , there is a significant causal relationship between menarche and vaginal (IVW, OR: 1.5108, 95% CI: 1.1474–2.0930, P  = 0.0043; MR-Egger, OR: 2.5575, 95% CI: 1.7664–9.6580, P  = 0.0013). The MR‐Egger intercepts showed no evidence for significant directional pleiotropy ( P intercept  = 0.1714 for menarche), suggesting that there is no directional pleiotropy in our two-sample MR. In two-way analyses, the vaginal-menarche causal model was not significant, suggesting no reverse causality (Figure S4). This evidence suggests that early stimulation of the vagina by high levels of estrogen (early menarche) may lead to the formation of a more effective barrier against external pathogens, thereby reducing the risk of vaginitis in women.

2-sample MR analysis. A Scatterplot of the causal effect of menarche on vaginitis. The slope of the line indicates the magnitude of the causal relationship. B Funnel plot. C Forest plot of MR analysis representing causal estimates of menarche on vaginitis. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple-instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. D LOO analyses. The black dots represent one IVW and the red dots represent estimates using all IVs. Horizontal lines indicate 95% confidence intervals

Causal relationship between age at menopause and estradiol and vaginitis

As age increases, a significant decrease in estrogen can lead to a decrease in vaginal lubrication, leading to vaginal atrophy, which is the most common problem during menopause [ 15 ]. Therefore, MR analysis was also performed for the causal effect between age at menopause and estradiol and vaginitis. The OR [95% CI] obtained by the IVW algorithm was 0.9872[0.9786–0.9959], the β value was − 0.0129, and the test P value was 0.0041. The Cochran Q test showed that there was no heterogeneity among the selected SNPs ( Q _pval = 0.9024). Due to the limited number of SNPs used, the funnel chart of 2-sample MR analysis was not drawn. Inversely, increased estrogen may delay the onset of menopause (Fig.  4 A–C). However, age at menopause had no significant effect on estradiol (Figure S5). Furthermore, neither age at menopause–vaginitis nor vaginitis–age at menopause causal models were significant in two-way analysis, suggesting no causal relationship between the two (Figures S6 and S7). SD increases in age of menopause, 1.001274 (95% CI 0.9689439–1.034682, P  = 0.9394043) for vaginitis; SD increases in age of menopause, 0.9180858 (95% CI 0.05868318–12.84376, P  = 0.9394043) for estradiol. Figure  4 D shows the schematic diagram of the main analysis results of this study.

Two-sample MR analysis. A Scatterplot of the causal effect of estradiol on age at menopause. The slope of the line indicates the magnitude of the causal relationship. B Funnel plot. C Forest plot of MR analysis representing causal estimates of estradiol on age at menopause. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple-instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. D Schematic diagram of the main analysis results of this study

Atrophic vaginitis (AV) is especially common in postmenopausal women with reduced ovarian estrogen production, and hypoestrogen (atrophic) individuals may have symptoms of vaginitis [ 21 ]. Currently, a variety of estrogen therapeutic agents are available, and their safety and efficacy in the treatment of AV have been evaluated in randomized controlled clinical trials [ 8 ]. MR analysis did not provide direct evidence of a causal relationship between estradiol and vaginitis. However, there was a reciprocal causal relationship between estradiol and age at menarche, and there was also a significant causal effect between menarche and vaginitis. This result suggests that when estrogen levels drop, it will lead to a delay in the age of menarche, and a late age of menarche may increase the risk of vaginitis, which may due to the fact that the vagina is stimulated by estrogen late and does not form a strong protective barrier. In conclusion, this study indirectly supports the association between increased estrogen and decreased risk of vaginitis, which is consistent with the results of clinical trials, suggesting that vaginitis may benefit from hormone therapy or replacement therapy [ 5 ].

Menarche usually occurs between the ages of 12 and 13, as the ovaries begin to produce estrogen, and the increase in these hormones prompts the thickening and preparation of the lining of the uterus, which eventually leads to menstrual cramps [ 13 , 22 ]. Age at menarche and age at natural menopause are affected by various environmental factors [ 23 ]. For age at menarche, the number of environmental covariates that could affect it was significantly smaller than age at natural menopause. For example, smoking, alcohol consumption, and duration of breastfeeding were identified as important factors affecting age at natural menopause [ 24 ]. Because they usually occur in adulthood, they probably do not affect age at menarche. However, the age of menarche is closely related to estrogen [ 25 ]. Multiple regression analysis revealed that age at menarche was inversely associated with E1 and E2 levels [ 26 ]. There are mouse data that suggest that circulating estradiol is critical for pubertal development [ 27 ]. Follicular phase serum estradiol levels are approximately twofold higher in women with early menarche compared with women with menarche at or after age 13 [ 28 ]. Estrogen excess is associated with early menstruation [ 29 ]. Although genome-wide association studies have reported associations between hundreds of genetic variants and age at menarche, studies have shown that specific genetic loci differentially influence age at menarche in healthy girls [ 30 ]. Recent genetic studies have identified rs10835638, a locus that is significantly associated with estradiol and causes menstrual cycle prolongation [ 31 ]. A later age at menarche may imply a later rise in estrogen levels, which may lead to physiological changes and poor defense against inflammation in the vaginal mucosa that make it more susceptible to infection, thereby increasing the risk of vaginitis [ 32 , 33 ]. However, this relationship is complex and is influenced by the interaction of multiple factors.

Late menarche and early menopause have been shown to be associated with poor health outcomes later in life [ 34 , 35 , 36 ]. Menopause is an important stage in a woman's life that affects women in many ways. The clinical symptoms of menopause are the result of the gradual depletion of oocytes in the ovary, resulting in a continuous decline in estrogen levels in the body [ 37 ]. As estrogen levels decline after menopause, vaginal tissue becomes atrophic (thinner, dry, and shrunken) [ 38 ]. Ovarian estrogen production begins to decline 1 or 2 years before menopause and reaches a stable nadir about 2 years after the last menstrual period. Serum concentrations of estradiol and estrone (the major circulating estrogens) are significantly lower compared to levels in women of reproductive age [ 39 ]. Menopausal women with moderate or severe symptoms are suitable candidates for hormone therapy in the absence of contraindications or other major comorbidities [ 40 ]. We present data showing that estradiol levels are inversely correlated with age at menopause and that increased estrogen may contribute to a later age at menopause. This finding is consistent with known physiological knowledge, as estradiol is one of the important hormones in the female reproductive system and plays a key role in ovarian function and reproductive cycle regulation. Again, this result is consistent with genetic evidence that lowering the T allele (rs10835638) causes delayed menopause [ 31 ]. However, we found no significant genetic causal relationship between age at menopause and vaginitis, which means that the age of menopause itself is not directly genetically related to the occurrence of vaginitis. Vaginitis may be affected by other factors, such as environmental factors, personal lifestyle, etc. In summary, the above results show that the longer the female reproductive tract is stimulated by estrogen, the stronger the defense ability formed and the lower the prevalence of vaginitis. Therefore, understanding the relationship between estrogen and vaginal health can help develop a more personalized treatment plan that is tailored to a patient's estrogen levels and other factors.

Previous analyses have shown that Lactobacillus has been widely used as one of the alternatives to routine antimicrobial treatment against vaginal pathogens for the prevention of chronic vaginitis and restoration of flora balance [ 41 , 42 ]. The data we presented showed that Lactobacillus was negatively correlated with vaginitis, but the difference was not significant. In addition, our analysis revealed an association between HPV16, HPV18 and estradiol and vaginitis, again not showing a significant causal effect. HPV16 and HPV18, as the two most common high-risk HPV subtypes, are highly associated with 70% of cervical cancers [ 41 ]. However, as far as current clinical research is concerned, the correlation between HPV and vaginitis and estradiol has not been fully explored [ 43 ].

A limitation that has to be mentioned in this study is that due to the limited number of SNPs used, there is a certain degree of heterogeneity among SNPs, which may affect our understanding of the causal effect between estrogen and vaginitis. Due to limitations of the dataset, we were unable to obtain detailed definitions of the phenotypes in each GWAS. In addition, lack of SNP data for vaginal microbes, we had to use data from gut microbes for MR analysis.

Limited by the diversity of detection methods and sample types, existing studies are heterogeneous. There is no consistent conclusion yet on the factors through which estradiol may affect vaginitis. In summary, our study provides supportive evidence for the treatment of vaginitis with estradiol from the perspective of genetic variation. While there is no direct causal relationship between estradiol and the risk of vaginitis, this study is the first to suggest that higher levels of estradiol may be linked to an earlier age of menarche. This could potentially result in a longer exposure of the vagina to estrogen, leading to the development of a stronger defense barrier and potentially reducing the risk of vaginitis. However, this process was not significantly associated with age at menopause, HPV, and Lactobacillus . Our study provides useful information for understanding the pathogenesis of vaginitis and provides important clues for future research directions in this field.

Availability of data and materials

The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request.

Vaginitis in Nonpregnant Patients. ACOG practice bulletin, number 215. Obstet Gynecol. 2020;135(1):e1–17.

Article   Google Scholar  

Buck ES, Lukas VA, Rubin RS. Effective prevention of recurrent UTIs with vaginal estrogen: pearls for a urological approach to genitourinary syndrome of menopause. Urology. 2021;151:31–6.

Article   PubMed   Google Scholar  

Safary M, et al. Comparison of the effects of fenugreek vaginal cream and ultra low- dose estrogen on atrophic vaginitis. Curr Drug Deliv. 2020;17(9):815–22.

Article   CAS   PubMed   Google Scholar  

Zhu YX, et al. Health-related quality of life as measured with the Short-Form 36 (SF-36) questionnaire in patients with recurrent vulvovaginal candidiasis. Health Qual Life Outcomes. 2016;14:65.

Article   PubMed   PubMed Central   Google Scholar  

Crandall CJ. Treatment of vulvovaginal atrophy. JAMA. 2019. https://doi.org/10.1001/jama.2019.15100 .

Yuldasheva F, Lapasova Z, Ismailova K. Algorithm of local therapy of vaginitis in postmenopausal women. Int Bull Med Sci Clin Res. 2023;3(1):89–95.

Google Scholar  

Cardozo L, et al. Meta-analysis of estrogen therapy in the management of urogenital atrophy in postmenopausal women: second report of the Hormones and Urogenital Therapy Committee. Obstet Gynecol. 1998;92(4 Pt 2):722–7.

CAS   PubMed   Google Scholar  

Lynch C. Vaginal estrogen therapy for the treatment of atrophic vaginitis. J Womens Health (Larchmt). 2009;18(10):1595–606.

Anahtar MN, et al. Cervicovaginal microbiota and reproductive health: the virtue of simplicity. Cell Host Microbe. 2018;23(2):159–68.

Smith BC, et al. Distinct ecological niche of anal, oral, and cervical mucosal microbiomes in adolescent women. Yale J Biol Med. 2016;89(3):277–84.

PubMed   PubMed Central   Google Scholar  

Turovskiy Y, Sutyak Noll K, Chikindas ML. The aetiology of bacterial vaginosis. J Appl Microbiol. 2011;110(5):1105–28.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Lebeau A, et al. HPV infection alters vaginal microbiome through down-regulating host mucosal innate peptides used by Lactobacilli as amino acid sources. Nat Commun. 2022;13(1):1076.

Karapanou O, Papadimitriou A. Determinants of menarche. Reprod Biol Endocrinol. 2010;8:115.

Barros BS, et al. ERICA: age at menarche and its association with nutritional status. J Pediatr (Rio J). 2019;95(1):106–11.

Pickar JH. Emerging therapies for postmenopausal vaginal atrophy. Maturitas. 2013;75(1):3–6.

Lawlor DA, et al. Mendelian randomization: using genes as instruments for making causal inferences in epidemiology. Stat Med. 2008;27(8):1133–63.

Davies NM, Holmes MV, Davey Smith G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ. 2018;362:k601.

Hemani G, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018. https://doi.org/10.7554/eLife.34408 .

Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017;32(5):377–89.

Bowden J, DaveySmith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512–25.

Donders G, et al. Ultra-low-dose estriol and Lactobacillus acidophilus vaginal tablets (Gynoflor®) for vaginal atrophy in postmenopausal breast cancer patients on aromatase inhibitors: pharmacokinetic, safety, and efficacy phase I clinical study. Breast Cancer Res Treat. 2014;145:371–9.

Omona K. Bleeding. In: Fibroids. 2020, IntechOpen.

Dvornyk V, et al. Predictive factors for age at menopause in Caucasian females. Maturitas. 2006;54(1):19–26.

Lu Y, et al. TNFRSF11A and TNFSF11 are associated with age at menarche and natural menopause in white women. Menopause. 2010;17(5):1048–54.

Ma H, et al. Reproductive factors and breast cancer risk according to joint estrogen and progesterone receptor status: a meta-analysis of epidemiological studies. Breast Cancer Res. 2006;8(4):R43.

Macmahon B, et al. Age at menarche, urine estrogens and breast cancer risk. Int J Cancer. 1982;30(4):427–31.

Clarkson J, et al. Postnatal development of an estradiol-kisspeptin positive feedback mechanism implicated in puberty onset. Endocrinology. 2009;150(7):3214–20.

Bernstein L, et al. Age at menarche and estrogen concentrations of adult women. Cancer Causes Control. 1991;2(4):221–5.

Song SS, Kang S, Park S. Association of estrogen-related polygenetic risk scores with breast cancer and interactions with alcohol intake, early menarche, and Nulligravida. Asian Pac J Cancer Prev. 2022;23(1):13–24.

Busch AS, et al. Differential impact of genetic loci on age at thelarche and menarche in healthy girls. J Clin Endocrinol Metab. 2018;103(1):228–34.

Ruth KS, et al. Genetic evidence that lower circulating FSH levels lengthen menstrual cycle, increase age at menopause and impact female reproductive health. Hum Reprod. 2016;31(2):473–81.

Shen J, et al. Effects of low dose estrogen therapy on the vaginal microbiomes of women with atrophic vaginitis. Sci Rep. 2016;6:24380.

Caussade A. The vulva during childhood. Vulvar disease: breaking the myths, 2019: p. 65–72.

Mishra GD, et al. Early life circumstances and their impact on menarche and menopause. Womens Health (Lond). 2009;5(2):175–90.

Wernli KJ, et al. Menstrual and reproductive factors in relation to risk of endometrial cancer in Chinese women. Cancer Causes Control. 2006;17(7):949–55.

Sturdee DW, Panay N. Recommendations for the management of postmenopausal vaginal atrophy. Climacteric. 2010;13(6):509–22.

Vigneswaran K, Hamoda H. Hormone replacement therapy—Current recommendations. Best Pract Res Clin Obstet Gynaecol. 2022;81:8–21.

Kelley C. Estrogen and its effect on vaginal atrophy in post-menopausal women. Urol Nurs. 2007;27(1):40–5.

PubMed   Google Scholar  

Henderson VW. Cognitive changes after menopause: influence of estrogen. Clin Obstet Gynecol. 2008;51(3):618–26.

Flores VA, Pal L, Manson JE. Hormone therapy in menopause: concepts, controversies, and approach to treatment. Endocr Rev. 2021;42(6):720–52.

Ma B, Forney LJ, Ravel J. Vaginal microbiome: rethinking health and disease. Annu Rev Microbiol. 2012;66:371–89.

Chee WJY, Chew SY, Than LTL. Vaginal microbiota and the potential of Lactobacillus derivatives in maintaining vaginal health. Microb Cell Fact. 2020;19(1):203.

Ramakrishnan S, Partricia S, Mathan G. Overview of high-risk HPV’s 16 and 18 infected cervical cancer: pathogenesis to prevention. Biomed Pharmacother. 2015;70:103–10.

Download references

Acknowledgements

Not applicable.

This work was supported by the General project of Shanghai Natural Science Foundation (18ZR1423900) and Shanghai Science and Technology Commission project (21D2312200).

Author information

Authors and affiliations.

Department of Gynecology and Obstetrics, Ruijin Hospital Affiliated to Shanghai Jiaotong University School of Medcine, 197 Ruijiner Road, Shanghai, 200003, China

Xiaosheng Xu, Yan Liu, Weiwei Feng & Jian Shen

You can also search for this author in PubMed   Google Scholar

Contributions

Study design: Jian Shen; analysis and interpretation of data: Xiaosheng Xu and Jian Shen; drafting of the manuscript: Jian Shen; critical revision of the manuscript for important intellectual content: Weiwei Feng and Yan Liu.

Corresponding authors

Correspondence to Weiwei Feng or Jian Shen .

Ethics declarations

Ethics approval and consent to participate, consent for publication, competing interests.

There are no conflicts of interest.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

40001_2024_1914_moesm1_esm.docx.

Supplementary Material 1. Figure S1: 2-sample MR analysis. (A) Scatterplot of the causal effect of estradiol on vaginitis. The slope of the line indicates the magnitude of the causal relationship. (B) Forest plot of MR analysis representing causal estimates of estradiol on vaginitis. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple-instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. (C) LOO analyses. The black dots represent one IVW and the red dots represent estimates using all IVs. Horizontal lines indicate 95% confidence intervals. Figure S2: 2-sample MR analysis. (A) Scatterplot of the causal effect of vaginitis on estradiol. The slope of the line indicates the magnitude of the causal relationship. (B) Forest plot of MR analysis representing causal estimates of vaginitis on estradiol. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple-instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. (C) LOO analyses. The black dots represent one IVW and the red dots represent estimates using all IVs. Horizontal lines indicate 95% confidence intervals. Figure S3: 2-sample MR analysis. (A) Scatterplot of the causal effect of menarche on estradiol. The slope of the line indicates the magnitude of the causal relationship. (B) Funnel plot. (C) Forest plot of MR analysis representing causal estimates of menarche on estradiol. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple-instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. (D) LOO analyses. The black dots represent one IVW and the red dots represent estimates using all IVs. Horizontal lines indicate 95% confidence intervals. (E) The 1SD value of the age at menarche. Figure S4: 2-sample MR analysis. (A) Scatterplot of the causal effect of vaginitis on menarche. The slope of the line indicates the magnitude of the causal relationship. (B) Forest plot of MR analysis representing causal estimates of vaginitis on menarche. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple-instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. (C) LOO analyses. The black dots represent one IVW and the red dots represent estimates using all IVs. Horizontal lines indicate 95% confidence intervals. Figure S5: 2-sample MR analysis. (A) Scatterplot of the causal effect of menarche on estradiol. The slope of the line indicates the magnitude of the causal relationship. (B) Forest plot of MR analysis representing causal estimates of menarche on estradiol. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple-instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. (C) LOO analyses. The black dots represent one IVW and the red dots represent estimates using all IVs. Horizontal lines indicate 95% confidence intervals. Figure S6: 2-sample MR analysis. (A) Scatterplot of the causal effect of age at menopause on vaginitis. The slope of the line indicates the magnitude of the causal relationship. (B) Forest plot of MR analysis representing causal estimates of age at menopause on vaginitis. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple-instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. (C) LOO analyses. The black dots represent one IVW and the red dots represent estimates using all IVs. Horizontal lines indicate 95% confidence intervals. Figure S7: 2-sample MR analysis. (A) Scatterplot of the causal effect of vaginitis on age at menopause. The slope of the line indicates the magnitude of the causal relationship. (B) Forest plot of MR analysis representing causal estimates of vaginitis on age at menopause. The circles next to each SNP indicate causal estimates for each IV, respectively, and the lowest two circles show multiple-instrument MR analysis using Egger regression and inverse-variance weighted methods. Horizontal lines denote 95% CIs. (C) LOO analyses. The black dots represent one IVW and the red dots represent estimates using all IVs. Horizontal lines indicate 95% confidence intervals.

Supplementary Material 2.

Supplementary material 3., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Xu, X., Liu, Y., Feng, W. et al. Strong evidence supports the use of estradiol therapy for the treatment of vaginal inflammation: a two-way Mendelian randomization study. Eur J Med Res 29 , 339 (2024). https://doi.org/10.1186/s40001-024-01914-4

Download citation

Received : 27 October 2023

Accepted : 31 May 2024

Published : 18 June 2024

DOI : https://doi.org/10.1186/s40001-024-01914-4

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Age at menarche
• Lactobacillus
• Mendelian randomization

European Journal of Medical Research

ISSN: 2047-783X

• Submission enquiries: [email protected]
• General enquiries: [email protected]