Global trends in dysmenorrhea research: A bibliometric analysis informing pain research collaboration and clinical translation.

Bibliometric analysis has emerged as an essential tool for examining the developmental context of academic disciplines, particularly within the realm of interdisciplinary research. This method quantitatively elucidates the evolution of knowledge structures and collaborative models, thereby offering researchers a strategic global perspective. [ 1 ] In recent years, the integration of network analysis and natural language processing (NLP) technologies has paved a novel pathway for large-scale literature mining. [ 2 ] Through the construction of topic co-occurrence networks [ 3 ] and author collaboration networks, [ 4 ] researchers can identify patterns of knowledge clustering and cross-institutional collaboration within the field. Furthermore, keyword timing analysis [ 5 ] and topic evolution mapping [ 6 ] facilitate the tracking of research hotspot migration patterns. Dysmenorrhea, a highly prevalent condition impacting 45% to 95% of women of reproductive age globally, [ 7 ] has witnessed a shift in research focus from the traditional examination of myometrial spasm mechanisms [ 8 ] to the investigation of the complex neuro-immune-endocrine regulatory network. [ 9 ] Over the past decade, there has been a 22.3% annual increase in the number of published studies. [ 10 ] Nevertheless, this surge in literature has introduced challenges such as fragmented knowledge and inefficient collaboration. Specifically, there exists a notable gap between fundamental research and clinical application, [ 11 ] and research efforts are predominantly concentrated in high-income countries, leading to significant biases in the analysis of epidemiological and sociocultural factors. Furthermore, existing reviews are constrained by manual synthesis methods and face difficulties in quantifying the structural deficiencies of international collaboration networks. [ 12 ] To achieve this objective, the present study develops an innovative framework that integrates multilevel econometric models, utilizing the R language software package “Bibliometrics” [ 13 ] (version 4.3.0; available at: https://www.bibliometrix.org ) to conduct a comprehensive bibliometric analysis of the literature. Specific tasks include constructing a global distribution network for “literature mutation analysis” and performing a detailed screening of periodicals, authors, citations, keywords, institutions, and countries, alongside a thorough analysis of co-occurrence network data. Subsequently, the data were exported in the format of a “PubMed Export File,” which contains the complete metadata from the PubMed database. [ 14 ] Data analysis was conducted using R language version 4.3.0 in conjunction with the bibliometrics software package. Preliminary data interpretation was facilitated through the use of the “biblioAnalysis()” [ 15 ] command and the “summary()” [ 16 ] function embedded within the software package. To establish a quantitative foundation for optimizing resource allocation and formulating prioritized research sequences – particularly in addressing the deficiencies in collaborative efforts within materials science for the development of analgesic patches [ 17 ] – this study proposes key policy intervention strategies.

Conceptualization: Yuying Zhang, Jinwen Sima, Xinyao Li, Shijia Lu, Yifei Chen, Minghua Dong. Data curation: Yuying Zhang, Jinwen Sima, Xinyao Li, Shijia Lu, Yifei Chen, Minghua Dong. Formal analysis: Yuying Zhang, Jinwen Sima, Xinyao Li, Shijia Lu, Yifei Chen, Minghua Dong. Funding acquisition: Yuying Zhang. Visualization: Yuying Zhang, Jinwen Sima. Writing – original draft: Yuying Zhang, Jinwen Sima, Xinyao Li, Shijia Lu, Yifei Chen, Minghua Dong. Writing – review & editing: Yuying Zhang, Jinwen Sima, Xinyao Li, Minghua Dong.

This study employs a multilevel bibliometric framework to elucidate the knowledge structure and evolutionary trends within the field of dysmenorrhea research through quantitative analysis. Data were sourced from the PubMed database, encompassing literature on dysmenorrhea from 2010 to 2025. The data processing was conducted using the “bibliometrix” software package (version 4.3.0) [ 18 ] in the R programming language. The specific methodology is detailed as follows. This study is grounded in a systematic literature review utilizing the PubMed database, employing a comprehensive search strategy to synthesize global scholarly contributions in the domain of dysmenorrhea research. The retrieval formula incorporates the subject term “dysmenorrhea” alongside its semantically related terms, including key medical terminology such as “primary dysmenorrhea,” “menstrual pain,” and “pelvic pain.” The scope of the literature is confined to original research articles, reviews, and clinical trials, with filtering criteria set to ARTICLE/REVIEW/CLINICAL TRIAL, spanning the years 2010 to 2025. This approach initially yielded 1528 literature records, comprising 903 research articles, 128 reviews, 46 case reports, 39 clinical trials, and 412 other types of documents. The data export adheres to the PubMed standard XML format protocol, [ 19 ] ensuring the complete retention of essential metadata fields, including title, author, affiliated institution, country, journal, keywords, abstract, citation network, publication year, and digital object identifier (DOI). Utilizing the “bibliometrix” software package in R (version 4.3.0), the original XML file is initially transformed into a structured data frame via the “convert2PubMed()” function, [ 20 ] facilitating the automated extraction of bibliometric fields. Subsequently, normalization of institution names is conducted, and any missing institutional information is supplemented through cross-validation with the Scopus database using the DOI identifier. [ 21 ] To resolve data redundancy, the Levenshtein similarity algorithm [ 22 ] (with a threshold of < 5) is employed to compare titles with DOI identifiers. As a result, 103 duplicate documents were eliminated, culminating in a final dataset comprising 1425 high-quality documents (Fig. 1 ). Comprehensive statistical analysis of literature metrics data indicators. This figure presents the core descriptive bibliometric indicators for the final dataset of 1425 high-quality valid dysmenorrhea research documents (retained after duplicate removal, data normalization, and preprocessing) retrieved from PubMed spanning 2010 to 2025, including 337 peer-reviewed academic journals publishing eligible studies, 5846 unique contributing authors, 216 distinct keywords extracted from titles and abstracts (note that “AB” in the original chart is a typographical artifact), an overall global co-authorship collaboration index of 0.72797 (which differs from the cross-country international collaboration intensity of 0.23 reported in the results section), 8573 total cited references across all included documents, and an average of 7.509 citations per document (note that the “%” symbol in the original chart is a formatting error as citation count is a non-percentage metric. This study developed 3 distinct types of core networks grounded in bibliometric principles to examine the global collaboration model and knowledge structure within dysmenorrhea research. All network analyses were conducted utilizing the “biblioNetwork()” function within the bibliometrix package. [ 23 ] Initially, the keyword co-occurrence network identified high-frequency topic words by employing a minimum word frequency threshold of 15. The correlation strength between terms (Jaccard similarity > 0.2) was calculated using the co-occurrence matrix generation algorithm, [ 24 ] and the Fruchterman-Reingold layout algorithm was employed to visualize topic clustering. [ 25 ] Subsequently, the author-institution collaboration network was established based on co-authorship relationships, with a minimum collaboration frequency requirement of 2, indicating that at least 2 papers must be co-published. Lastly, the literature co-citation network extracted 124 high-impact publications with a citation frequency of ≥ 20 using the “citationNetwork()” function, [ 26 ] and the knowledge evolution trajectory was illustrated through timeline visualization. [ 27 ] In the interim, statistical techniques, including factor analysis, [ 28 ] were employed to reduce dimensionality and categorize variables within the literature dataset. Through such methods, the primary factors or elements within the dataset were identified, facilitating the classification and analysis of literature or authors. The positioning of words on the Word Map indicates their scores or contributions to the 2 principal components, Dim1 and Dim2. Typically, words in closer proximity exhibit a higher degree of similarity, whereas those more distantly spaced demonstrate lower similarity. All analyses were conducted using R version 4.3.2, and the source code repository has been made publicly accessible. The dynamic topic evolution tracking technology was employed to construct the topic evolution pathway within the domain of dysmenorrhea research utilizing the “thematicEvolution()” function from the bibliometrics software package. [ 29 ] This model segmented the period from 2010 to 2025 into 4 successive time intervals. By examining changes in the modular structure of the keyword co-occurrence network, it computed topic similarity (Jaccard coefficient ≥ 0.4) and derivation strength ( S value ≥ 0.7), [ 30 ] and validated the robustness of the evolution pathway through Monte Carlo simulation (1000 iterations, P  < .01). A Gini coefficient exceeding 0.68 was considered indicative of high concentration, [ 31 ] thereby elucidating the evolution trajectory of the core research topics and the emergence patterns of new hotspots. All analyses were conducted using R version 4.3.2, and the source code library was made publicly accessible.

Through a comprehensive, multidimensional bibliometric analysis encompassing keyword distribution, topic evolution, academic influence, and cooperative networks, the critical issue of the disconnect between “pain mechanism” and “clinical translation” in dysmenorrhea research has been distinctly identified. The analysis of keyword and thematic evolution reveals that terms associated with pain mechanisms, such as “dysmenorrhea” and “primary dysmenorrhea,” are central to the research focus, exhibiting prominence with a font size 3 to 4 times larger than other terms. These terms emphasize foundational investigations, including prostaglandin regulation and neuro-immune-endocrine networks. In contrast, clinically translational themes like “quality of life” and “pain management” have been historically marginalized, only appearing in statistical analyses post-2020, with a frequency <50% of that of the core terms. This disparity is further corroborated by the topological structure of the thematic co-occurrence network. The investigation into pain mechanisms constitutes a central cluster within the “dysmenorrhea → pain → endometriosis” framework, exhibiting a centrality of 0.71. In contrast, the theme of clinical transformation requires traversing 3 levels of transitions to establish associations, with a minimum edge weight of 1.2, and its evolutionary trajectory remains isolated, as indicated by a Jaccard association strength of <0.1. An analysis of academic influence and collaborative networks has identified the structural root cause of this disconnection: the intensity of mutual citation between basic and clinical research clusters is merely 0.03, and cross-cluster connections account for only 12%, significantly lower than the intra-cluster average of 68%. China leads in output, with 38.7% of corresponding authors, yet this output predominantly circulates within domestic journals, creating a closed cycle. The research on pain mechanisms is notably deficient in clinical validation stages, resulting in a “high output − low impact” paradox, with the average annual citation rate projected to decline to zero by 2025. Countries such as Iran have attained a peak citation rate of 2.4 by concentrating on mechanical-translational research, including areas like nonsteroidal anti-inflammatory drug resistance and adolescent pain intervention. However, the global cooperation network exhibits structural deficiencies. In China, the domestic cooperation density (0.48) is 8 times greater than that of international cooperation. Furthermore, collaboration across institutions and borders remains insufficient, hindering the effective integration of knowledge on pain mechanisms with clinical requirements. To address the disconnect between “pain mechanisms and clinical translation,” a 3-tiered intervention strategy should be adopted. In the short term (1–3 years), efforts should concentrate on disrupting the existing closed cycle. This can be achieved by promoting the submission of research on pain mechanisms to clinical translational journals, with the aim of having international translational journals comprise over 40% of these submissions. Additionally, priority should be given to funding research focused on “mechanism-oriented clinical interventions,” such as personalized treatments via the prostaglandin pathway and studies verifying the neuroimmune mechanisms underlying the efficacy of acupuncture. In the medium term (3–5 years), efforts should focus on reconstructing the knowledge transmission channel and mandating the integration of basic research projects within the clinical transformation module. This includes the establishment of a joint laboratory dedicated to “pain mechanism − clinical transformation” and the promotion of cross-cluster collaboration. The objective is to enhance the correlation between the mechanism and transformation topics to exceed 0.5. Additionally, a reform of the evaluation system is necessary, ensuring that translational indicators, such as clinical trial outcomes and improvements in patient health, are given equal importance alongside the outputs of basic research. In the long term (5–10 years), the development of a global transformation collaboration network is envisaged. This will leverage the South-South cooperation model, combining China’s expertise in mechanism research with Iran’s experience in transformation. The collaboration will extend to underserved regions such as Africa and the Middle East, with the aim of increasing the node degree of the Iranian network to over 10. Concurrently, a global dysmenorrhea research data platform will be established to consolidate mechanism data, clinical trial results, and real-world outcomes. Expedite the translation of pain mechanism research into clinical guidelines and treatment protocols to effectively address the issue of disconnection.

The authors would like to express their gratitude to the PubMed database for providing the relevant literature data. They also acknowledge the developers of the R language and the bibliometrix software package for offering powerful tools for bibliometric analysis. Additionally, thanks are extended to all researchers whose work is cited in this study, as their contributions laid the foundation for this bibliometric analysis.