Research scope and analytical framework

This study examines endocrinology-related literature indexed in the Web of Science (WoS) Core Collection from 2000 to 2023, analyzing 19,127 articles retrieved on October 17, 2024. Network analysis is used to elucidate collaboration patterns, identify leading researchers, and highlight prominent institutions in the field. The analysis tracks the evolution of international cooperation, identifies periods of intensified collaboration, and explores the impact of these relationships on research productivity and innovation in endocrinology.

The WoS data reveals notable trends in publication volume. Publications increased sharply in 2009, peaked around 2012, declined temporarily in 2014, and resurged starting in 2019 (Figure 1). Advancements in technologies such as next-generation sequencing and enhanced diagnostic tools likely drove the initial surge. The growing prevalence of diabetes and obesity heightened research interest further. The 2019 resurgence reflects progress in therapeutic areas, such as immunotherapy and gene therapy, alongside increased research addressing the interplay between endocrine function and infectious diseases, particularly focusing on COVID-19’s implications for endocrine health (11).

• Data Collection: This study analyzed co- authorship networks based on endocrinology- related articles indexed in the WoS Core Collection, covering the period from 2000 to 2023. A total of 19,127 articles published during this period (as of October 2024) were collected. These articles were retrieved using the search topic “Endocrinology” to capture relevant publications, providing a comprehensive dataset to evaluate collaborative patterns in this field.

• Analytical tools and environment: The analysis was conducted using the Python programming language (version 3.10.5) within the PyCharm Integrated Development Environment (IDE), version 2022.1.3. Python’s robust libraries for network analysis, including NetworkX, provided the computational tools necessary to calculate both macro-level and micro-level network metrics in the co-authorship networks.

• Network analysis methodology: The network analysis aims to examine the structure and collaborative dynamics of the co-authorship networks in endocrinology, with a focus on the following macro-level and micro-level metrics:

Macro-level metrics

Network density: Calculated as the ratio of the actual number of edges to the maximum possible number of edges in the network, this metric provides insight into the overall connectivity of the network (12).

Figure 1: Trend of the number of papers published in the Web of Science Core Collection

Source: Own elaboration

• Clustering coefficient: This metric measures the extent to which nodes in the network tend to cluster together, indicating the degree of collaboration in research groups (13).

• Components: The analysis identified and counted connected components within the network to understand the number and size of independent subgroups, illustrating the fragmentation or integration in research collaboration (14).

• Average distance: The average shortest path length between nodes in the network was calculated, reflecting how closely connected researchers are across the entire network

.

Micro-level metrics

• Degree centrality: This metric represents the number of edges each node has, measuring each researcher’s direct connections and their importance within the network (12).

• Closeness centrality: Closeness centrality measures how close each node is to all other nodes in the network, indicating the efficiency with which a researcher can reach others (13).

• Betweenness centrality: This metric evaluates the extent to which a node lies on the shortest path between other nodes, assessing the influence of a researcher as a bridge between different parts of the network (14)

.

These metrics, grounded in network science methodology, enable an in-depth analysis of the collaborative structure and highlight key researchers in the field of endocrinology, allowing us to track the evolution of collaborative patterns from 2000 to 2023.

2000-2009 Analysis of endocrinology research networks

During the 2000-2009 period, the network density of endocrinology research was calculated at 0.00044, indicating a sparse network with limited collaboration among researchers (Table 1). Despite this low density, the average clustering coefficient was high (0.9126), showing that research clusters, though loosely connected, tended to have strong internal collaborations (Figure 2). The network comprised 2,192 distinct components, suggesting fragmented collaboration structures (Table 1). The average distance was infinite, reflecting disconnected components (15).

Figure 2. : Top 20 endocrinology researcher network from 2000 to 2009 Note. This figure visualizes the co-authorship network of the top 20 researchers in endocrinology between 2000 and 2009Node sizes are proportional to degree centrality, and edges represent collaborative relationships.

In terms of centrality measures, the top contributors by degree centrality included Oberfield SE (0.0049), Levine LS (0.0045), and Savage MO (0.0044), indicating their frequent collaborations (Table 2). By closeness centrality, Savage MO (0.0113) and Grüters A (0.0110) demonstrated high accessibility across collaborative pathways (Table 3). Betweenness centrality revealed Grossman AB (0.0004) and Savage MO (0.0004) as critical intermediaries, enhancing connectivity among disparate research groups (Table 4).

2010-2019 Analysis of endocrinology research networks

From 2010 to 2019, network density decreased slightly to 0.00025, and the network became more fragmented, with 3,550 components (Table 1).

However, the clustering coefficient remained high at 0.9129 (Table 1), maintaining strong localized collaboration patterns (Figure 3). The average distance metric remained infinite, indicating continued separation among components (15).

Figure 3: Top 20 endocrinology researcher network from 2010 to 2019 Note. This figure visualizes the co-authorship network of the top 20 researchers in endocrinology between 2010 and 2019. Node sizes are proportional to degree centrality, and edges represent collaborative relationships.

Source: Own elaboration

This period saw Yildiz M. (0.0047) and Boysan SN (0.0047) as top nodes in degree centrality, highlighting their wide-reaching collaborations (Table 2). In closeness centrality, Murad M. Hassan (0.0813) and Pasquali Renato (0.0796) emerged as highly accessible researchers (Table 3). Murad M. Hassan (0.0275) led in betweenness centrality, followed by Pasquali Renato (0.0136), underscoring their role as central connectors facilitating cross- group collaborations (Table 4).

Global research distribution and collaboration dynamics

The WoS data demonstrates concentrated endocrinology research output from the United States, with substantial contributions from the United Kingdom, Germany, France, and other European countries (Figure 4). China, Japan, and India lead Asian contributions, while Brazil remains the primary contributor in Latin America.

This geographic distribution highlights both global commitment to addressing endocrine disorders and regional differences in research focus.

Figure 4: Top 20 endocrinology researcher network from 2020 to 2023 Note. This figure visualizes the co-authorship network of the top 20 researchers in endocrinology between 2020 and 2023. Node sizes are proportional to degree centrality, and edges represent collaborative relationships.

Source: Own elaboration

The analysis explores these international trends in depth, focusing on key collaboration indicators to evaluate network structure evolution. Co-authorship ties reveal patterns of cross-border collaboration and the influence of key researchers in fostering transnational research partnerships. These insights illuminate how endocrinology research networks adapt to emerging challenges and opportunities.

2020-2023 Analysis of endocrinology research networks

The 2020-2023 period demonstrated an increase in network density to 0.00037, alongside a high clustering coefficient of 0.9329 (Table 1), which reflected an increasingly connected network structure with robust collaborative clusters (Figure 5). The network consisted of 2,712 components, suggesting a consolidation of collaborations compared to the previous period (Figure 5). The average distance remained infinite, consistent with the presence of segmented network structure (15).

Figure 5: Total number of papers published by country in the Web of Science Core Collection (2000 - 2023)

Source: Own elaboration

Top nodes by degree centrality in this period included Sahin Mustafa (0.0059) and Polak Michel (0.0045), indicating their extensive research collaborations (Table 2). Polak Michel (0.0725) and Umpierrez Guillermo E. (0.0722) scored highest in closeness centrality, making them central figures in collaboration pathways (Table 3). Betweenness centrality values were highest for Ji Linong (0.0368) and Kosiborod Mikhail (0.0154), emphasizing their roles as pivotal links across the research network (Table 4).

Summary of endocrinology research networks analysis (2000 - 2023)

Across the years analyzed, the endocrinology research network evolved from a fragmented structure with low density to a more interconnected network with increased collaboration and clustering. Researchers like Savage MO (University College London, UK), Murad M. Hassan (Mayo Clinic, USA), and Ji Linong (Peking University, China) consistently maintained central positions, indicating their significant influence in fostering cross-group collaborations. The high clustering coefficients across periods highlight the presence of strongly connected collaborative clusters, which underpin the advancement of endocrinology research.

Evolution of network structure and collaborative patterns

The endocrinology research network exhibited distinct characteristics across three time periods analyzed. From 2000 to 2009, the network displayed sparse connectivity, with low density and high fragmentation, as evidenced by multiple components and an infinite average distance. The high clustering coefficient during this period indicates that researchers primarily collaborated within specific clusters, typically bounded by institutional or geographic boundaries. This pattern reflected the contemporary research funding landscape, which often confined collaboration within regional contexts.

During the 2010-2019 period, the network exhibited increased fragmentation while maintaining strong clustering patterns. The decreased network density combined with elevated clustering coefficients suggests that researchers concentrated their collaborative efforts within small, cohesive research groups. Notable researchers such as Murad M. Hassan and Pasquali Renato served as crucial intermediaries during this period, connecting disparate research groups and promoting international collaboration. This evolution coincided with significant advances in precision medicine and next-generation sequencing, which encouraged specialized research partnerships.

From 2020 to 2023, network density showed modest improvement, indicating partial consolidation of collaborative structures. The persistent high clustering coefficient suggests that while segmentation remained, these segments maintained strong internal cohesion. Researchers, including Mustafa Sahin, emerged as central network figures, promoting transnational collaboration. This period also corresponded with an increase in interdisciplinary research addressing global health challenges, particularly diabetes, metabolic syndrome, and COVID-19’s impact on endocrine health.

Impact of central figures and institutions

Throughout these periods, researchers with high centrality measures served as collaboration hubs. Figures such as Savage MO, Murad M. Hassan, and Ji Linong consistently demonstrated leadership by contributing substantially to their research domains while facilitating connections across geographic and institutional boundaries. Their affiliations with prestigious institutions, like University College London, the Mayo Clinic, and Peking University, underscore the influential role of established research centers in driving the field’s advancement.

Researchers with high betweenness centrality, particularly Ji Linong and Mikhail Kosiborod, functioned as critical network bridges, enabling knowledge and resource transfer between otherwise disconnected groups. This bridging function proves especially valuable in endocrinology, where diverse research areas-from genetic studies to lifestyle disease management-demand multidisciplinary approaches.

Implications and future directions

The analysis highlights the critical role of collaborative networks in advancing endocrinology research. While the consistently high clustering coefficient indicates strong intra-group collaboration, network fragmentation suggests opportunities for enhanced interdisciplinary and international partnerships. The persistence of regional clusters, particularly among researchers in North America, Europe, and Asia, reflects varying research priorities based on local health challenges. North American and European researchers often prioritize lifestyle-related diseases, while Asian researchers contribute significantly to the study of genetic and environmentally influenced disorders.

This research provides a framework for understanding collaborative dynamics in endocrinology and identifies opportunities for bridging network gaps. Future initiatives should focus on fostering cross-regional and interdisciplinary collaborations to address emerging global health challenges more effectively. National governments, healthcare institutions, and academic societies-particularly associations of internal medicine and endocrinology-must establish stronger partnerships to promote interdisciplinary approaches. These institutional stakeholders should coordinate their efforts to facilitate joint research programs capable of addressing global health challenges that require worldwide coordination and resources. Such systematic collaboration among policymakers, healthcare providers, and academic organizations will create a more robust framework for tackling complex endocrine disorders that transcend national boundaries.

The findings from this network analysis provide valuable insights into the collaborative landscape of endocrinology research, highlighting the roles of key figures, institutions, and geographic regions in shaping the field. These insights suggest strategic directions for strengthening global efforts in endocrine disorder research and treatment.

Significance of the study

This mapping of collaborative dynamics within endocrinology research provides valuable insights on how scientific cooperation shapes the field’s growth. The identification of influential researchers, leading research institutions, and key collaboration clusters contributes to understanding the factors driving endocrinology’s rapid evolution. Network structure examination over time offers a model for assessing research networks in other medical fields, revealing how collaboration dynamics influence scientific discovery’s pace and direction. This analysis emphasizes international collaboration’s pivotal role in advancing scientific innovation and addressing endocrine disorders’ growing complexity, paving the way for improved clinical outcomes worldwide.

The study presents a comprehensive view of the collaborative framework within endocrinology, demonstrating how collective research efforts worldwide contribute to understanding and addressing pressing medical challenges.