Global Research Landscape of Mitochondrial Dysfunction in Sepsis: A Comprehensive Bibliometric and Visualization Analysis (2000–2025) Authors: Shanshan Wang, Chong Meng, Hongjun Bian, Xin Zhang, Minllin Cong, Jinlin Ren, Jingzhao Sun, Haiyang Zhang, Dongxing Liu Published: 2026 ABSTRACT ObjectiveSepsis-associated mitochondrial dysfunction represents a central but incompletely understood mechanism underlying organ failure and mortality in critically ill patients. Despite a growing body of literature, no comprehensive bibliometric analysis has systematically mapped the global research landscape, knowledge structure, and emerging frontiers of this field. This study aimed to quantify publication trends, identify key contributors, delineate thematic hotspots, and characterize the evolving knowledge base in sepsis and mitochondria research over the past 25 years.MethodsA systematic search of the Web of Science Core Collection was conducted using a predefined search strategy combining sepsis- and mitochondria-related terms, restricted to publications from January 1, 2000, to December 31, 2025. After sequential exclusion of retracted publications, non-article/review document types, and non-English records, a final dataset of 3545 publications was obtained. Bibliometric analyses were performed using the bibliometrix package in R, VOSviewer (version 1.6.20), and CiteSpace (version 6.2.R4). Analyses included annual publication trends, geographic and institutional distribution, Bradford’s Law journal analysis, Lotka’s Law author productivity, co-citation and bibliographic coupling network analysis, thematic mapping, and keyword burst detection.ResultsAnnual publication output increased more than nine-fold from 2000 to 2025, with a marked acceleration following the 2016 Sepsis-3 consensus definition. China led in publication volume (n = 1350; 38.1%), while the United States demonstrated the highest citation impact (52,732 total citations; 67.61 average citations per publication). Bradford’s Law analysis identified 30 core journals, led by Shock, Frontiers in Immunology, and Critical Care Medicine. Lotka’s Law analysis confirmed that 74.94% of contributing authors published a single article. The most cited publication was Zhang Q et al (Nature, 2010; 3080 citations). Thematic map analysis identified mitochondrial dysfunction/acute kidney injury, oxidative stress/reactive oxygen species, and inflammation/mitochondrial DNA as Motor Themes. Keyword burst analysis revealed a temporal evolution from early nitric oxide and cytochrome c research toward current hotspots including NLRP3 inflammasome activation, sepsis-associated encephalopathy, and sepsis-induced cardiomyopathy.ConclusionThis comprehensive bibliometric analysis provides a systematic overview of 25 years of research at the intersection of sepsis and mitochondrial biology. The field is experiencing sustained growth, with current research frontiers focused on organ-specific mitochondrial injury, innate immune regulation, and neurological complications. Introduction Sepsis is a life-threatening medical emergency defined as organ dysfunction caused by a dysregulated host response to infection, and remains one of the most pressing challenges in contemporary critical care medicine. Despite decades of research and advances in supportive care, sepsis and septic shock continue to impose a major global health burden, with high rates of morbidity and mortality persisting across healthcare systems worldwide. According to the Global Burden of Disease Study 2017, sepsis was responsible for an estimated 48.9 million incident cases and 11.0 million deaths globally, accounting for approximately 19.7% of all global deaths.1 The clinical management of sepsis is guided by the Surviving Sepsis Campaign international guidelines, most recently updated in 2026, which provide evidence-based recommendations for hemodynamic resuscitation, infection source control, antimicrobial therapy, and organ support.2 Despite adherence to these guidelines, in-hospital mortality for sepsis in high-income countries remains 15–25%, rising to 30–40% in septic shock, with substantially higher mortality rates reported in low- and middle-income countries.1 The disproportionate burden of sepsis in resource-limited settings underscores the urgent need for deeper mechanistic understanding and the identification of novel therapeutic targets that could be translated into accessible clinical interventions. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3), published in 2016, fundamentally reframed sepsis as a syndrome of dysregulated host response leading to organ dysfunction, thereby redirecting scientific attention toward the molecular and cellular mechanisms of organ failure rather than the systemic inflammatory response syndrome per se.3 This conceptual shift has catalyzed a new wave of mechanistic research, with mitochondrial biology emerging as one of the most actively investigated pathophysiological domains. Among the diverse pathophysiological mechanisms implicated in sepsis-induced organ dysfunction, mitochondrial dysfunction has emerged as a particularly central and potentially reversible contributor. Mitochondria are not merely the primary site of cellular ATP generation through oxidative phosphorylation; they are increasingly recognized as dynamic signaling organelles that integrate metabolic, immunological, and cell death pathways in response to infectious and inflammatory stimuli. In the context of sepsis, mitochondrial injury is initiated through multiple converging mechanisms, including the overproduction of reactive oxygen and nitrogen species that directly inhibit electron transport chain complexes I and IV, impaired mitophagy leading to the accumulation of dysfunctional organelles, dysregulation of mitochondrial fission and fusion dynamics mediated by dynamin-related protein 1, and mitochondrial membrane permeabilization triggering intrinsic apoptotic cascades.4 The clinical relevance of these mechanisms was definitively established by Brealey et al (Lancet, 2002), who demonstrated a direct correlation between the degree of mitochondrial dysfunction in skeletal muscle biopsies and both the severity and outcome of human septic shock,5 providing the first compelling evidence that mitochondrial bioenergetic failure is not merely an epiphenomenon but a mechanistic driver of organ failure. Subsequent research has broadened this framework considerably. Mitochondrial dysfunction is now recognized as a unifying pathophysiological thread connecting diverse forms of sepsis-induced organ injury, including acute kidney injury, cardiomyopathy, encephalopathy, and diaphragmatic failure, with organ-specific differences in mitochondrial abundance, metabolic substrate dependence, and susceptibility to injury collectively shaping the clinical pattern of multi-organ dysfunction.6 The discovery that damaged mitochondria release danger-associated molecular patterns—including mitochondrial DNA and formyl peptides—into the circulation further established mitochondria as active participants in the amplification of systemic inflammation, rather than passive victims of it.7 More recently, the concept of metabolic hibernation or cytopathic hypoxia has been proposed to explain the paradox of organ dysfunction persisting despite apparently adequate oxygen delivery, with mitochondrial dysfunction and consequent ATP depletion driving a state of adaptive hypometabolism that, while potentially protective in the short term, contributes to prolonged organ dysfunction when unresolved.4 Collectively, these mechanistic insights have generated substantial scientific and therapeutic interest in mitochondria as both diagnostic biomarkers and pharmacological targets in sepsis. Bibliometric analysis has emerged as a powerful and widely adopted methodological approach for systematically mapping the intellectual landscape of rapidly evolving biomedical research fields. In the field of sepsis research, bibliometric analyses have been published with increasing frequency over the past decade, addressing a range of subfields including septic shock,8 sepsis-associated acute kidney injury,9 sepsis-associated encephalopathy,10 sepsis immunosuppression,11 endothelial cell dysfunction in sepsis,12 and artificial intelligence applications in sepsis diagnosis and management.13 However, a critical and conspicuous gap exists in the current bibliometric literature: despite the central importance of mitochondrial dysfunction in sepsis pathophysiology, no comprehensive bibliometric analysis has to date systematically characterized the global research landscape at this intersection. The absence of a systematic bibliometric overview means that researchers, clinicians, and funding agencies currently lack a quantitative roadmap for understanding how this field has evolved, which research themes are most central, and where scientific investment would be most productively directed. To address this critical knowledge gap, the present study aimed to conduct the first comprehensive bibliometric and visualization analysis of global research on sepsis and mitochondrial dysfunction, encompassing all publications indexed in the Web of Science Core Collection between January 2000 and December 2025. Specifically, we sought to: (1) characterize temporal trends in publication output and citation impact; (2) identify the leading contributing countries, institutions, and authors; (3) delineate the core journal landscape using Bradford’s Law; (4) characterize author productivity distribution using Lotka’s Law; (5) map the knowledge structure through co-citation and bibliographic coupling network analysis; (6) identify research themes and their developmental trajectories through thematic mapping; and (7) delineate temporal research hotspots and emerging frontiers through keyword burst detection. The findings of this study are intended to provide researchers, clinicians, and funding agencies with a quantitative and evidence-based roadmap of the current state and future directions of this rapidly evolving field. Methods Data Source and Search Strategy All bibliometric data were retrieved from the Web of Science Core Collection (WoSCC), which was selected as the sole database due to its comprehensive coverage of high-quality peer-reviewed literature and its compatibility with downstream bibliometric analysis tools. A systematic search was conducted on April 27, 2026, using the following Topic field (TS) search query: TS=(“sepsis” OR “septicemia” OR “septic shock”) AND TS=(“mitochondria” OR “mitochondrial” OR “mitophagy” OR “mitochondrial dysfunction” OR “mitochondrial biogenesis” OR “mitochondrial dynamics”) AND PY=(2000–2025) NOT DT=(“Retracted Publication”). To minimize the impact of database indexing lag on the completeness of recent literature, the search was intentionally conducted in April 2026, approximately four months after the end of the study period (December 31, 2025), allowing sufficient time for the indexing of publications from the final year of analysis. In addition, 31 Early Access records whose Final Publication Year had been assigned to 2026 at the time of retrieval were identified, retained, and recoded to 2025 to further enhance the completeness of the most recent literature. Inclusion and Exclusion Criteria Records were included if they met all of the following criteria: (1) published as original articles or review articles; (2) written in English; and (3) indexed within the defined search period (2000–2025). Records were excluded if they were: (1) retracted publications; (2) non-English language publications; or (3) document types other than articles or reviews. Thirty-one records retrieved as Early Access publications whose Final Publication Year was assigned to 2026 were retained and recoded to 2025. The complete screening and selection process is illustrated in Figure 1. Figure 1Flow diagram of the literature screening and selection process. Records were retrieved from the Web of Science Core Collection using a predefined search strategy covering the period 2000–2025. Sequential filtering steps excluded retracted publications (n = 18), non-article/review document types (n = 299), and non-English records (n = 46), yielding a final analytical dataset of 3545 publications. Thirty-one Early Access articles indexed within the 2025 search window whose Final Publication Year was assigned to 2026 were retained and recoded to 2025.A flowchart of literature screening from database search to final dataset. Flow diagram of the literature screening and selection process. Records were retrieved from the Web of Science Core Collection using a predefined search strategy covering the period 2000–2025. Sequential filtering steps excluded retracted publications (n = 18), non-article/review document types (n = 299), and non-English records (n = 46), yielding a final analytical dataset of 3545 publications. Thirty-one Early Access articles indexed within the 2025 search window whose Final Publication Year was assigned to 2026 were retained and recoded to 2025. Data Extraction and Processing Full bibliographic records were exported from WoSCC in plain text format in batches of up to 500 records and subsequently merged into a single dataset. Country name standardization was performed prior to analysis: “Peoples R China” was recoded as “China”; “Taiwan” was merged into “China” in institutional address fields; regional entities of the United Kingdom were unified as “United Kingdom”; and “Turkiye” was harmonized as “Turkey”. All data processing was performed in R (version 4.3.1). These standardization procedures were performed to ensure consistency in geographic attribution and to align with the country and region classification conventions adopted by the Web of Science database and widely used in previous bibliometric studies. Specifically, the constituent regions of the United Kingdom (England, Scotland, Wales, and Northern Ireland) were consolidated under “United Kingdom” in accordance with their status as a single sovereign country, and address records were harmonized accordingly. These procedures were applied solely for the purpose of standardized bibliometric aggregation and do not reflect any political position on the part of the authors. Bibliometric Analysis Bibliometric analyses were performed using the bibliometrix package (version 4.1.4) in R. Basic descriptive statistics were calculated, including annual publication output, total citation counts, and country-, institution-, and author-level productivity metrics. Bradford’s Law analysis was conducted to identify core journals. Lotka’s Law analysis was performed to characterize the distribution of scientific productivity among authors. A thematic map was constructed based on author keyword co-occurrence data to identify and classify research themes into four quadrants: Motor Themes, Niche Themes, Basic Themes, and Emerging or Declining Themes. Descriptive statistics were used to summarize publication counts, citation frequencies, and productivity metrics. For Lotka’s Law analysis, the goodness-of-fit between the observed author productivity distribution and the theoretical Lotka distribution was assessed using the Kolmogorov–Smirnov test, with the Lotka exponent estimated by the bibliometrix package. For Bradford’s Law, journals were partitioned into three zones each accounting for approximately one-third of the total articles. Collaboration and co-occurrence network clustering was performed using the Louvain community detection algorithm, and the total link strength was calculated to quantify the intensity of collaborative and co-citation relationships. All analyses were descriptive in nature, consistent with the methodological framework of bibliometric research, and no inferential hypothesis testing was applied beyond the goodness-of-fit assessment described above. Network Visualization Collaboration and co-occurrence networks were visualized using VOSviewer (version 1.6.20). Three distinct network analyses were performed: (1) an international collaboration network based on country-level co-authorship, with a minimum publication threshold of five documents per country; (2) an institutional collaboration network based on organization-level co-authorship, with a minimum publication threshold of 25 documents per institution; and (3) a document co-citation network, with a minimum co-citation threshold of 30 citations per reference, and a bibliographic coupling network of documents with a minimum citation threshold of 250 citations per document. Clustering in all networks was performed using the Louvain algorithm. Keyword Burst Detection Keyword burst detection analysis was conducted using CiteSpace (version 6.2.R4). The analysis covered the period 2000–2025, with a time slice of one year per slice. Node type was set to keyword, and the Pathfinder pruning algorithm was applied. The top 25 keywords with the strongest citation burst values were identified and visualized. Annual Trend Analysis Annual publication trends were analyzed using R with ggplot2 (version 3.4.4). A locally estimated scatterplot smoothing (LOESS) curve was fitted to the annual publication data. The year 2016 was designated as a key temporal landmark corresponding to the publication of the Sepsis-3 consensus definition, dividing the study period into two phases: Phase I (Pre-Sepsis-3, 2000–2015) and Phase II (Post-Sepsis-3, 2016–2025). Results Overall Publication Output The systematic search of the WoSCC yielded 3908 initial records. Following sequential exclusion of 18 retracted publications, 299 non-article/review document types, and 46 non-English records, a final analytical dataset of 3545 publications was obtained (Figure 1). These publications were contributed by authors from 62 countries, affiliated with 2080 unique institutions, and published across 973 distinct journals. The dataset comprised 2757 original articles and 695 review articles, with a total citation count exceeding 120,000 at the time of data retrieval. Temporal Trends in Publication Output Annual publication output demonstrated a sustained and marked growth trajectory over the 25-year study period (Figure 2). The number of annual publications increased from fewer than 50 in 2000 to 464 in 2025, representing a greater than nine-fold increase. The LOESS trend curve revealed two distinct growth phases. During Phase I (2000–2015, Pre-Sepsis-3), publication growth was gradual and relatively modest. A marked acceleration in publication output was observed following 2016, coinciding with the introduction of the Sepsis-3 consensus definition. Total annual citations peaked around 2018–2020, with a subsequent apparent decline attributable to citation lag in more recently published articles. Figure 2Annual publication output and cumulative citation trends in sepsis and mitochondria research from 2000 to 2025. Blue bars represent the number of publications per year; the dashed blue line indicates a locally estimated scatterplot smoothing (LOESS) trend. The red line depicts total annual citations. A vertical dotted line marks 2016, the year of the Sepsis-3 consensus definition, which serves as a temporal landmark dividing Phase I (Pre-Sepsis-3, 2000–2015) and Phase II (Post-Sepsis-3, 2016–2025). Citation counts for publications from 2020 onwards reflect incomplete accumulation attributable to citation lag and should not be interpreted as indicating reduced research impact.A combined bar and line chart displays annual publications, total citations and a LOESS trend from 2000 to 2025. The x-axis shows publication years, while the left y-axis indicates the number of publications (0-500) and the right y-axis shows total citations (0-10,000). A dotted line at 2016 marks Sepsis-3, dividing Phase I Pre-Sepsis-3 and Phase II Post-Sepsis-3. Publications rise from 20 in 2000 to 460 in 2025, with notable increases post-2016. Citations start at 1500 in 2000, peaking at 11,200 in 2022, then drop to 1500 by 2025. The LOESS trend shows a steady rise from 25 in 2000 to 440 in 2025.A combined bar and line chart showing annual publications and total citations by publication year. Annual publication output and cumulative citation trends in sepsis and mitochondria research from 2000 to 2025. Blue bars represent the number of publications per year; the dashed blue line indicates a locally estimated scatterplot smoothing (LOESS) trend. The red line depicts total annual citations. A vertical dotted line marks 2016, the year of the Sepsis-3 consensus definition, which serves as a temporal landmark dividing Phase I (Pre-Sepsis-3, 2000–2015) and Phase II (Post-Sepsis-3, 2016–2025). Citation counts for publications from 2020 onwards reflect incomplete accumulation attributable to citation lag and should not be interpreted as indicating reduced research impact. Geographic Distribution and International Collaboration Publications originated from 62 countries worldwide (Table 1 and Figure 3). China was the most prolific contributor (n = 1350, 38.1%), followed by the United States (n = 780, 22.0%), Germany (n = 131, 3.7%), the United Kingdom (n = 125, 3.5%), and Brazil (n = 104, 2.9%). The United States demonstrated the highest total citation count (52,732) and average citations per publication (67.61). The United Kingdom exhibited the highest average citation rate among the top 10 countries (105.06 citations per publication). The international collaboration network revealed five principal research clusters.Table 1Top 10 Countries by Publication Output in Sepsis and Mitochondria Research (2000–2025)RankCountryDocumentsCitationsAverage Citations1China135032,59724.152USA78052,73267.613Germany131604546.154United Kingdom12513,133105.065Brazil104360934.76Japan87300934.597France79411952.148Canada78424154.379Korea78511565.5810Netherlands64337452.72 Figure 3International collaboration network in sepsis and mitochondria research (2000–2025). Each node represents a country; node size is proportional to publication volume. Edges indicate co-authorship relationships between countries, and edge thickness reflects collaboration frequency. Colors denote distinct research clusters identified by the Louvain algorithm. Only countries with a minimum of five publications are displayed (n = 50).A bibliometric network map shows country nodes connected by curved lines, with node sizes and link thicknesses varying. USA and China are central, linked to many countries. Medium-sized nodes like the UK, Italy, Netherlands, Spain, Japan, Canada, Germany, France, Switzerland and Brazil form a dense web with numerous interconnections. European countries such as Belgium, Austria, Sweden, Portugal, Greece, Hungary, Slovenia, Turkey, Poland, Iran and Russia are tightly interlinked with central hubs. Smaller nodes like Australia, Israel, Ireland, Thailand, Mexico, South Korea, New Zealand, Egypt, Saudi Arabia, Singapore, India and South Africa are farther from the core with fewer connections. Countries like Czech Republic, Serbia, Indonesia, Slovakia, Denmark, Argentina and Romania connect to the main network via long links. No axis, scale, or units are shown. VOSviewer is labeled at the bottom left.A diagram showing a bibliometric country co authorship network map with labeled nodes and links. Table 1Top 10 Countries by Publication Output in Sepsis and Mitochondria Research (2000–2025)RankCountryDocumentsCitationsAverage Citations1China135032,59724.152USA78052,73267.613Germany131604546.154United Kingdom12513,133105.065Brazil104360934.76Japan87300934.597France79411952.148Canada78424154.379Korea78511565.5810Netherlands64337452.72 Top 10 Countries by Publication Output in Sepsis and Mitochondria Research (2000–2025) International collaboration network in sepsis and mitochondria research (2000–2025). Each node represents a country; node size is proportional to publication volume. Edges indicate co-authorship relationships between countries, and edge thickness reflects collaboration frequency. Colors denote distinct research clusters identified by the Louvain algorithm. Only countries with a minimum of five publications are displayed (n = 50). Institutional Collaboration Network Institutional analysis identified 2080 unique contributing organizations worldwide (Table 2 and Figure 4). Among institutions with a minimum of 25 publications, 33 met the threshold for network visualization. The Pennsylvania Commonwealth System of Higher Education led in total publication volume (n = 213), followed by Harvard University (n = 204), the University of Pennsylvania (n = 189), and the University of Pittsburgh (n = 186). Among Chinese institutions, Southern Medical University (n = 176), Central South University (n = 137), and Shanghai Jiao Tong University (n = 128) were the most productive. The network revealed two principal clusters: a Western cluster centered on North American and European research universities, and a Chinese cluster comprising leading medical universities.Table 2Top 10 Institutions by Publication Output in Sepsis and Mitochondria Research (2000–2025)RankInstitutionDocuments1Pennsylvania Commonwealth System of Higher Education (PCSHE)2132Harvard University2043University of Pennsylvania1894University of Pittsburgh1865Southern Medical University - China1766Harvard University Medical Affiliates1717University of London1638University College London1419Central South University13710Shanghai Jiao Tong University128 Figure 4Institutional collaboration network in sepsis and mitochondria research (2000–2025). Each node represents an institution; node size is proportional to publication volume. Edges indicate co-authorship relationships, and edge thickness reflects collaboration frequency. Colors denote distinct institutional clusters identified by the Louvain algorithm. Only institutions with a minimum of 25 publications are displayed (n = 33).A network diagram shows institutional co-authorship links as circular nodes connected by curved lines. Most nodes form a dense web, while a smaller chain links back to the main web via long arcs. Visible node labels include Shanghai Jiao Tong Univ, Fudan Univ, Zhejiang Univ, Wenzhou Med Univ, China Med Univ, Capital Med Univ, Cent South Univ, Huazhong Univ Sci and Technol, Wuhan Univ, Univ Pittsburgh, Harvard Univ, Ohio State Univ, Univ Penn, Southern Med Univ, Guangzhou Med Univ, Sun Yat Sen Univ, Sichuan Univ, Tianjin Med Univ, Nanjing Univ, Nanjing Med Univ, Soochow Univ and Chongqing Med Univ. Node sizes vary, indicating different magnitudes and curve thickness varies, indicating connection strengths. Nodes form distinct clusters with many intra-cluster and some cross-cluster links. A small label, VOSviewer, is near the bottom edge.A diagram showing an institutional co authorship network with sized nodes and weighted connecting lines. Table 2Top 10 Institutions by Publication Output in Sepsis and Mitochondria Research (2000–2025)RankInstitutionDocuments1Pennsylvania Commonwealth System of Higher Education (PCSHE)2132Harvard University2043University of Pennsylvania1894University of Pittsburgh1865Southern Medical University - China1766Harvard University Medical Affiliates1717University of London1638University College London1419Central South University13710Shanghai Jiao Tong University128 Top 10 Institutions by Publication Output in Sepsis and Mitochondria Research (2000–2025) Institutional collaboration network in sepsis and mitochondria research (2000–2025). Each node represents an institution; node size is proportional to publication volume. Edges indicate co-authorship relationships, and edge thickness reflects collaboration frequency. Colors denote distinct institutional clusters identified by the Louvain algorithm. Only institutions with a minimum of 25 publications are displayed (n = 33). Author Productivity and Lotka’s Law A total of 15,487 unique authors contributed to the dataset (Table 3). Zhang Y was the most prolific author (n = 62 publications), followed by Singer M (n = 59) and Wang Y (n = 55). Singer M demonstrated the highest citation impact, with 5467 total citations and an average of 92.66 citations per publication. Lotka’s Law analysis revealed that 74.94% of all contributing authors published only a single article in this domain (Table 4). The fitted Lotka exponent was 2.469, approximating the theoretical value of 2 that characterizes the inverse square law of scientific productivity. A Kolmogorov–Smirnov goodness-of-fit test was performed to compare the observed author productivity distribution with the theoretical Lotka distribution (D = 0.44, p = 0.006), and the overall distribution pattern was broadly consistent with Lotka’s Law, with a small number of highly prolific authors accounting for a disproportionate share of the total output. As shown in Table 3, the majority of the most productive authors were affiliated with Chinese institutions, with the notable exception of Singer M (University of London, United Kingdom), who also demonstrated the highest citation impact among the leading authors, further underscoring the disparity between publication volume and citation influence discussed below.Table 3Top 10 Most Productive Authors in Sepsis and Mitochondria Research (2000–2025), with Their Primary Country and Institutional AffiliationRankAuthorDocumentsCitationsAverage CitationsCountryPrimary Institution1Zhang Y62206133.24ChinaTianjin Medical University2Singer M59546792.66United KingdomUniversity of London3Wang Y55267748.67ChinaZhejiang University4Yang Y46135229.39ChinaNorthwest University Xi’an5Zhang J39135634.77ChinaSouthern Medical University6Li Y37101827.51ChinaChang Gung Memorial Hospital7Wang J37110829.95ChinaWuhan University8Li XY33131839.94ChinaWuhan University9Wang L3161119.71ChinaNanchang University10Chen Y283715132.68ChinaChang Gung Memorial HospitalNote: Average citations were calculated as total citations divided by the number of publications. Country and institutional affiliations were assigned based on the most frequent affiliation recorded across each author’s publications, following the country standardization procedures described in the Methods. Table 4Lotka’s Law Analysis of Author Productivity Distribution in Sepsis and Mitochondria Research (2000–2025)Documents (n)AuthorsObserved (%)Expected (%)111,60674.9474.942201212.9918.7437414.788.3343752.424.6852331.503.0061280.832.087990.641.538580.371.179580.370.9310490.320.75Note: Lotka’s exponent (β) = 2.469; Kolmogorov–Smirnov goodness-of-fit test, D = 0.44, p = 0.006. The observed distribution follows the general pattern of Lotka’s Law, in which the proportion of authors producing n publications is approximately proportional to 1/n2. Table 3Top 10 Most Productive Authors in Sepsis and Mitochondria Research (2000–2025), with Their Primary Country and Institutional AffiliationRankAuthorDocumentsCitationsAverage CitationsCountryPrimary Institution1Zhang Y62206133.24ChinaTianjin Medical University2Singer M59546792.66United KingdomUniversity of London3Wang Y55267748.67ChinaZhejiang University4Yang Y46135229.39ChinaNorthwest University Xi’an5Zhang J39135634.77ChinaSouthern Medical University6Li Y37101827.51ChinaChang Gung Memorial Hospital7Wang J37110829.95ChinaWuhan University8Li XY33131839.94ChinaWuhan University9Wang L3161119.71ChinaNanchang University10Chen Y283715132.68ChinaChang Gung Memorial HospitalNote: Average citations were calculated as total citations divided by the number of publications. Country and institutional affiliations were assigned based on the most frequent affiliation recorded across each author’s publications, following the country standardization procedures described in the Methods. Top 10 Most Productive Authors in Sepsis and Mitochondria Research (2000–2025), with Their Primary Country and Institutional Affiliation Note: Average citations were calculated as total citations divided by the number of publications. Country and institutional affiliations were assigned based on the most frequent affiliation recorded across each author’s publications, following the country standardization procedures described in the Methods. Table 4Lotka’s Law Analysis of Author Productivity Distribution in Sepsis and Mitochondria Research (2000–2025)Documents (n)AuthorsObserved (%)Expected (%)111,60674.9474.942201212.9918.7437414.788.3343752.424.6852331.503.0061280.832.087990.641.538580.371.179580.370.9310490.320.75Note: Lotka’s exponent (β) = 2.469; Kolmogorov–Smirnov goodness-of-fit test, D = 0.44, p = 0.006. The observed distribution follows the general pattern of Lotka’s Law, in which the proportion of authors producing n publications is approximately proportional to 1/n2. Lotka’s Law Analysis of Author Productivity Distribution in Sepsis and Mitochondria Research (2000–2025) Note: Lotka’s exponent (β) = 2.469; Kolmogorov–Smirnov goodness-of-fit test, D = 0.44, p = 0.006. The observed distribution follows the general pattern of Lotka’s Law, in which the proportion of authors producing n publications is approximately proportional to 1/n2. Journal Distribution and Bradford’s Law The 3545 publications were distributed across 973 unique journals (Table 5 and Figure 5). Bradford’s Law analysis stratified these journals into three zones of approximately equal article contribution. Zone 1 (core journals) comprised 30 journals accounting for 1176 articles (33.2%). The leading journal was Shock (n = 133), followed by Frontiers in Immunology (n = 84), Critical Care Medicine (n = 83), Critical Care (n = 72), and International Immunopharmacology (n = 66). Zone 2 comprised 154 journals (1203 articles), and Zone 3 encompassed 790 journals (1166 articles).Table 5Top 10 Journals Publishing Sepsis and Mitochondria Research with Bradford Zone Classification (2000–2025)RankJournalDocumentsBradford Zone1Shock133Zone 12Frontiers in Immunology84Zone 13Critical Care Medicine83Zone 14Critical Care72Zone 15International Immunopharmacology66Zone 16International Journal of Molecular Sciences63Zone 17Free Radical Biology and Medicine51Zone 18PLOS ONE49Zone 19Scientific Reports44Zone 110Frontiers in Pharmacology36Zone 1 Figure 5Bradford’s Law distribution of journal productivity in sepsis and mitochondria research (2000–2025). The curve depicts the cumulative number of articles published across journals ranked by productivity. Vertical dashed lines demarcate three Bradford zones of equal article contribution: Zone 1 (Core, n = 30 journals, 1176 articles), Zone 2 (n = 154 journals, 1203 articles), and Zone 3 (n = 790 journals, 1166 articles).A line graph showing cumulative number of articles by journal rank across Bradford Zone 1, Zone 2 and Zone 3. Table 5Top 10 Journals Publishing Sepsis and Mitochondria Research with Bradford Zone Classification (2000–2025)RankJournalDocumentsBradford Zone1Shock133Zone 12Frontiers in Immunology84Zone 13Critical Care Medicine83Zone 14Critical Care72Zone 15International Immunopharmacology66Zone 16International Journal of Molecular Sciences63Zone 17Free Radical Biology and Medicine51Zone 18PLOS ONE49Zone 19Scientific Reports44Zone 110Frontiers in Pharmacology36Zone 1 Top 10 Journals Publishing Sepsis and Mitochondria Research with Bradford Zone Classification (2000–2025) Bradford’s Law distribution of journal productivity in sepsis and mitochondria research (2000–2025). The curve depicts the cumulative number of articles published across journals ranked by productivity. Vertical dashed lines demarcate three Bradford zones of equal article contribution: Zone 1 (Core, n = 30 journals, 1176 articles), Zone 2 (n = 154 journals, 1203 articles), and Zone 3 (n = 790 journals, 1166 articles). Most Cited Publications The top 10 most cited publications are presented in Table 6. The most highly cited work was Zhang Q et al (Nature, 2010; 3080 citations; 181.18 citations/year), reporting the release of mitochondrial DNA as a damage-associated molecular pattern in trauma and sepsis. The second most cited publication was Nakahira K et al (Nature Immunology, 2011; 2591 citations), demonstrating the role of mitochondrial reactive oxygen species in NLRP3 inflammasome activation.Table 6Top 10 Most Cited References in Sepsis and Mitochondria Research (2000–2025)RankReferenceCitationsCitations/Year1Zhang Q, 2010, Nature3080181.182Nakahira K, 2011, Nat Immunol2591161.943Islam MN, 2012, Nat Med125283.474Kaplan MJ, 2012, J Immunol97665.075Yousefi S, 2008, Nat Med78541.326Murakami T, 2012, P Natl Acad Sci USA76851.27Hotchkiss RS, 2001, J Immunol75028.858Manne BK, 2020, Blood719102.719Lelubre C, 2018, Nat Rev Nephrol68375.8910Ince C, 2005, Crit Care64429.27 Table 6Top 10 Most Cited References in Sepsis and Mitochondria Research (2000–2025)RankReferenceCitationsCitations/Year1Zhang Q, 2010, Nature3080181.182Nakahira K, 2011, Nat Immunol2591161.943Islam MN, 2012, Nat Med125283.474Kaplan MJ, 2012, J Immunol97665.075Yousefi S, 2008, Nat Med78541.326Murakami T, 2012, P Natl Acad Sci USA76851.27Hotchkiss RS, 2001, J Immunol75028.858Manne BK, 2020, Blood719102.719Lelubre C, 2018, Nat Rev Nephrol68375.8910Ince C, 2005, Crit Care64429.27 Top 10 Most Cited References in Sepsis and Mitochondria Research (2000–2025) Keyword Burst Analysis Citation burst analysis identified 25 keywords with the strongest temporal burst signals (Figure 6). Early-phase bursts (2000–2015) were dominated by “nitric oxide” (burst strength 41.59), “nitric oxide synthase” (28.42), and “tumor necrosis factor” (17.93). The transitional phase (2005–2017) was characterized by “multiple organ failure” and “cytochrome c oxidase”. Recent-phase bursts (2021–2025) included “sepsis-associated encephalopathy” (14.96), “NLRP3 inflammasome” (12.21), “sepsis-induced cardiomyopathy” (12.32), and “macrophage polarization” (8.79). Figure 6Top 25 keywords with the strongest citation bursts in sepsis and mitochondria research (2000–2025), generated using CiteSpace. Each row represents a keyword; the red segment indicates the period of burst activity, and the blue/green segment represents the non-burst period. Burst strength values are shown for each keyword. Keywords are ranked by burst strength in descending order. The strongest early bursts were observed for “nitric oxide” (strength 41.59) and “nitric oxide synthase” (28.42) during 2000–2015, while the most recent bursts, including “sepsis-associated encephalopathy” (14.96) and “macrophage polarization” (8.79), extended through 2025, reflecting the temporal shift in research frontiers.A horizontal bar graph titled, Top 25 Keywords with the Strongest Citation Bursts. Left table columns read, Keywords, Year, Strength, Begin, End. A timeline at right is labeled 2000 minus 2026. X axis label, 2000 minus 2026, unit, year. Y axis label, Keywords, unit, keyword. Rows and values: nitric oxide, Year 2000, Strength 41.59, Begin 2000, End 2014. tumor necrosis factor, 2000, 17.93, 2000, 2012. endotoxin, 2000, 11.73, 2000, 2016. cytochrome c, 2000, 10.14, 2000, 2011. blood flow, 2000, 9.89, 2000, 2014. nitric oxide synthase, 2001, 28.42, 2001, 2015. peroxynitrite, 2001, 17.91, 2001, 2014. skeletal muscle, 2001, 17.36, 2001, 2011. complex i, 2001, 12.93, 2001, 2014. free radicals, 2001, 11.62, 2001, 2007. rat, 2001, 9.63, 2001, 2011. improves survival, 2001, 9.26, 2001, 2013. oxygen consumption, 2002, 11.28, 2002, 2015. gene expression, 2003, 9.2, 2003, 2014. organ failure, 2000, 9.86, 2004, 2013. necrosis factor alpha, 2004, 9.83, 2004, 2012. multiple organ failure, 2001, 12.46, 2005, 2017. cytochrome c oxidase, 2005, 10.58, 2005, 2015. in vivo, 2006, 13.69, 2006, 2016. nlrp3 inflammasome, 2014, 12.21, 2021, 2026. sepsis minus associated encephalopathy, 2011, 14.96, 2022, 2026. sepsis minus induced cardiomyopathy, 2022, 12.32, 2022, 2026. myocardial injury, 2022, 10.87, 2022, 2026. septic cardiomyopathy, 2015, 9.98, 2023, 2026. macrophage polarization, 2021, 8.79, 2023, 2026.A bar graph showing top 25 keywords with the strongest citation bursts over time. Top 25 keywords with the strongest citation bursts in sepsis and mitochondria research (2000–2025), generated using CiteSpace. Each row represents a keyword; the red segment indicates the period of burst activity, and the blue/green segment represents the non-burst period. Burst strength values are shown for each keyword. Keywords are ranked by burst strength in descending order. The strongest early bursts were observed for “nitric oxide” (strength 41.59) and “nitric oxide synthase” (28.42) during 2000–2015, while the most recent bursts, including “sepsis-associated encephalopathy” (14.96) and “macrophage polarization” (8.79), extended through 2025, reflecting the temporal shift in research frontiers. Thematic Map Analysis Thematic map analysis identified six thematic clusters across four quadrants (Figure 7). Three clusters were identified in Motor Themes: “sepsis/mitochondrial dysfunction/acute kidney injury”, “mitochondria/oxidative stress/reactive oxygen species”, and “inflammation/metabolism/mitochondrial DNA”. One Niche Theme cluster was identified: “sepsis-associated encephalopathy/neuroinflammation/microglia”. Two Emerging or Declining Themes clusters were identified: “apoptosis/lipopolysaccharide/rat” and “septic shock/critical illness/microcirculation”. No themes were identified in the Basic Themes quadrant. Figure 7Thematic map of research topics in sepsis and mitochondria research (2000–2025), generated using the bibliometrix package in R. Each bubble represents a thematic cluster; bubble size is proportional to keyword frequency. The horizontal axis (Centrality) reflects the degree of interaction between a theme and other themes; the vertical axis (Density) reflects the internal development of a theme. Four quadrants are defined: Motor Themes (high centrality, high density), Niche Themes (low centrality, high density), Emerging or Declining Themes (low centrality, low density), and Basic Themes (high centrality, low density). No themes were identified in the Basic Themes quadrant. The three Motor Theme clusters (sepsis/mitochondrial dysfunction/acute kidney injury; mitochondria/oxidative stress/reactive oxygen species; and inflammation/metabolism/mitochondrial DNA) occupied positions of both high centrality and high density, indicating well-developed themes central to the field. The single Niche Theme (sepsis-associated encephalopathy/neuroinflammation/microglia) exhibited high density but low centrality, representing a specialized subfield.A bubble scatter plot is divided into four quadrants by dashed lines. The top left quadrant is labeled ′Niche Themes,′ the top right ′Motor Themes,′ and the bottom left ′Emerging or Declining Themes.′ The horizontal axis represents ′Relevance degree (Centrality)′ and the vertical axis ′Development degree (Density),′ without numeric labels. Six bubbles are labeled: In the upper right, a large bubble is marked ′sepsis, mitochondrial dysfunction, acute kidney injury.′ Two overlapping bubbles near the upper center are labeled ′mitochondria, oxidative stress, reactive oxygen species′ and ′inflammation, metabolism, mitochondrial DNA.′ In the upper left, a bubble is labeled ′sepsis-associated encephalopathy, neuroinflammation, microglia.′ In the lower left, bubbles are labeled ′septic shock, critical illness, microcirculation′ and ′apoptosis, lipopolysaccharide, rat.′ A size legend at the lower right shows frequency with circles labeled 500, 1,500 and 3,000.A bubble scatter plot showing thematic clusters by centrality and density in sepsis and mitochondria research. Thematic map of research topics in sepsis and mitochondria research (2000–2025), generated using the bibliometrix package in R. Each bubble represents a thematic cluster; bubble size is proportional to keyword frequency. The horizontal axis (Centrality) reflects the degree of interaction between a theme and other themes; the vertical axis (Density) reflects the internal development of a theme. Four quadrants are defined: Motor Themes (high centrality, high density), Niche Themes (low centrality, high density), Emerging or Declining Themes (low centrality, low density), and Basic Themes (high centrality, low density). No themes were identified in the Basic Themes quadrant. The three Motor Theme clusters (sepsis/mitochondrial dysfunction/acute kidney injury; mitochondria/oxidative stress/reactive oxygen species; and inflammation/metabolism/mitochondrial DNA) occupied positions of both high centrality and high density, indicating well-developed themes central to the field. The single Niche Theme (sepsis-associated encephalopathy/neuroinflammation/microglia) exhibited high density but low centrality, representing a specialized subfield. Knowledge Structure: Co-Citation and Bibliographic Coupling Networks Document co-citation network analysis (87 references cited ≥30 times) identified three principal knowledge clusters anchored by the works of Brealey et al (Lancet, 2002), Singer et al (JAMA, 2016), and seminal immunological studies (Figure 8A). Bibliographic coupling network analysis (81 documents with ≥250 citations) revealed four current research frontiers focused on cardiovascular complications, renal complications, mitochondrial bioenergetics, and antioxidant therapeutics (Figure 8B). Figure 8Bibliometric network analysis of the knowledge structure in sepsis and mitochondria research (2000–2025). (A) Co-citation network of cited references. Each node represents a cited reference; node size reflects citation frequency. Edges indicate co-citation relationships, and colors denote distinct knowledge clusters identified by the Louvain algorithm. Only references cited ≥30 times are displayed (n = 87). (B) Bibliographic coupling network of documents. Each node represents a published article; node size reflects citation count. Edges indicate shared reference overlap between documents. Only documents with ≥250 citations are displayed (n = 81). Together, (A and B) reveal the foundational knowledge base and current research frontiers of the field, respectively. In (A) the co-citation network was anchored by the seminal works of Brealey et al (2002) and Singer et al (2016); in (B) the bibliographic coupling network delineated four contemporary research directions, encompassing cardiovascular complications, renal complications, mitochondrial bioenergetics, and antioxidant therapeutics.The image A showing a co-citation network of cited references. Text at the top reads, A. Nodes are labeled with author, year and journal, including singer, m. 2016 jama j am med and brealey, david 2002 lancet. Many curved connecting lines link nodes into several dense clusters, with a few large, central nodes connected to many others. The image B showing a bibliographic coupling network of documents. Text at the top reads, B. Nodes are labeled with author and year, including brealey (2002) and singer (2016). Curved connecting lines form multiple clusters, including one cluster centered around galluzzi (2011) and another around lebleu (2018). Several nodes appear larger and more connected than surrounding nodes, indicating higher citation count and stronger shared-reference overlap.Two-panel infographic of bibliometric networks on sepsis and mitochondria research. Bibliometric network analysis of the knowledge structure in sepsis and mitochondria research (2000–2025). (A) Co-citation network of cited references. Each node represents a cited reference; node size reflects citation frequency. Edges indicate co-citation relationships, and colors denote distinct knowledge clusters identified by the Louvain algorithm. Only references cited ≥30 times are displayed (n = 87). (B) Bibliographic coupling network of documents. Each node represents a published article; node size reflects citation count. Edges indicate shared reference overlap between documents. Only documents with ≥250 citations are displayed (n = 81). Together, (A and B) reveal the foundational knowledge base and current research frontiers of the field, respectively. In (A) the co-citation network was anchored by the seminal works of Brealey et al (2002) and Singer et al (2016); in (B) the bibliographic coupling network delineated four contemporary research directions, encompassing cardiovascular complications, renal complications, mitochondrial bioenergetics, and antioxidant therapeutics. Discussion Publication Growth Trends and the Impact of Sepsis-3 This bibliometric analysis of 3545 publications retrieved from the Web of Science Core Collection provides the first comprehensive mapping of global research trends at the intersection of sepsis and mitochondrial biology over a 25-year period (2000–2025). The sustained and accelerating growth in annual publication output reflects the growing recognition of sepsis as a global health emergency. According to the Global Burden of Disease Study, sepsis was responsible for approximately 11 million deaths worldwide in 2017, accounting for nearly 20% of all global mortality,1 a burden that has galvanized international research efforts across basic science, translational, and clinical domains. The identification of mitochondrial dysfunction as a central, potentially reversible mechanism underlying sepsis-induced organ failure has further amplified scientific interest in this area. The landmark study by Brealey et al (Lancet, 2002) provided the first direct human evidence linking mitochondrial dysfunction to the severity and outcome of septic shock,5 establishing a mechanistic framework that continues to underpin the field to this day. A particularly noteworthy observation is the marked inflection in publication output following 2016, coinciding with the publication of the Sepsis-3 consensus definition. Singer et al (JAMA, 2016) redefined sepsis as life-threatening organ dysfunction caused by a dysregulated host response to infection,3 explicitly placing organ dysfunction at the centre of the clinical definition. This paradigm shift appears to have catalyzed a wave of mechanistic research aimed at elucidating the molecular basis of organ dysfunction, with mitochondrial pathways emerging as a particularly active area of investigation. Similar post-Sepsis-3 publication surges have been reported in bibliometric analyses of sepsis-associated acute kidney injury and sepsis-associated encephalopathy, suggesting that the Sepsis-3 framework has broadly stimulated organ-specific mechanistic research across the critical care community. Geographic Distribution and Institutional Collaboration The geographic distribution of publications revealed that China was the most prolific contributor, accounting for 38.1% of the total dataset (n = 1350), followed by the United States (22.0%). This finding is consistent with the broader trajectory of Chinese biomedical research output, which has expanded dramatically over the past two decades driven by sustained government investment in research infrastructure.14 A similar pattern of Chinese dominance in publication volume has been consistently observed across other sepsis-related bibliometric analyses, including those focused on sepsis immunosuppression,11 septic intestinal injury,15 and artificial intelligence applications in sepsis,13 suggesting this reflects a field-wide phenomenon. The present analysis reveals an important qualitative disparity between publication volume and citation impact. Despite leading in publication output, China’s average citation rate (24.15) was substantially lower than that of the United States (67.61) and the United Kingdom (105.06). A similar pattern has been documented in the field of artificial intelligence and sepsis research.13 This divergence is likely multifactorial rather than attributable to any single cause. First, the relatively limited intensity of international collaboration among Chinese institutions may constrain citation impact, as internationally co-authored publications generally attract higher citation rates; the institutional network analysis in the present study indeed revealed that Chinese institutions formed a relatively self-contained cluster with sparse connections to Western research centers. Second, differences in the distribution of publications across high-impact Zone 1 journals may contribute, as a greater proportion of highly cited landmark studies, consensus definitions, and mechanistic discoveries have historically originated from institutions in the United States and the United Kingdom. Third, the rapid expansion of publication output may be accompanied by a degree of thematic homogenization, whereby a large volume of publications concentrates on similar topics, potentially diluting per-article citation impact. These factors collectively suggest that enhancing international collaboration and prioritizing high-impact, mechanistically novel research may help improve the citation influence of the substantial research output from Chinese institutions. Most Cited Publications and Knowledge Structure The most highly cited work was Zhang Q et al (Nature, 2010; 3080 citations), which demonstrated that mitochondria release DAMPs into the circulation following injury, activating a systemic inflammatory response.7 The second most cited publication was Nakahira K et al (Nature Immunology, 2011; 2591 citations), which demonstrated that mitochondrial reactive oxygen species are essential for NLRP3 inflammasome activation.16 The co-citation network identified three principal knowledge clusters. The first was anchored by Brealey et al (Lancet, 2002),5 establishing foundational evidence for mitochondrial dysfunction as a central mechanism in sepsis-induced organ failure. The second was centered on the Sepsis-3 definition paper. The third comprised immunological and molecular biology references related to mitochondrial signaling. A comparable tripartite knowledge architecture has been identified in bibliometric analyses of sepsis endothelial dysfunction.12 Thematic Map Analysis and Research Topic Architecture The largest Motor Theme cluster—comprising “sepsis”, “mitochondrial dysfunction”, and “acute kidney injury”—reflects the well-established pathophysiological link between mitochondrial bioenergetic failure and sepsis-induced renal injury. Mitochondrial dysfunction is now recognized as one of the central pathogenic mechanisms of SA-AKI, encompassing impaired oxidative phosphorylation, mitochondrial dynamics dysregulation, defective mitophagy, and excessive reactive oxygen species generation.17 The second Motor Theme cluster— “mitochondria”, “oxidative stress”, and “reactive oxygen species”—represents the canonical mechanistic pathway of mitochondrial injury in sepsis.18 The third Motor Theme cluster— “inflammation”, “metabolism”, and “mitochondrial DNA”—reflects the intersection of immunometabolism and mitochondrial danger signaling. During the hyperinflammatory phase of sepsis, energy is preferentially generated through glycolysis instead of oxidative phosphorylation, a phenomenon termed metabolic reprogramming,19 while mitochondrial DNA released into the circulation sustains systemic inflammation through pattern recognition receptor activation.7 The Niche Theme cluster— “sepsis-associated encephalopathy”, “neuroinflammation”, and “microglia”—occupies a position of high internal density but lower centrality. Neuroinflammation caused by hyperactivation of microglia is considered a key pathogenic factor, initiating a cascade including blood-brain barrier disruption, oxidative stress, and neuronal injury.20 Mitochondrial dysfunction mediated through dynamin-related protein 1 has been implicated in propagating blood-brain barrier impairment in septic encephalopathy.21 The isolation of this cluster from the central network suggests that the mitochondrial mechanisms underlying neurological complications of sepsis remain incompletely integrated into the broader field, representing a priority area for future investigation. Keyword Burst Analysis and Emerging Research Frontiers During the early burst phase (2000–2015), the dominant emerging keywords included “nitric oxide” (burst strength 41.59), “nitric oxide synthase” (28.42), “tumor necrosis factor” (17.93), and “cytochrome c” (10.14), reflecting the prevailing paradigm in which sepsis-associated mitochondrial injury was primarily understood through the lens of nitric oxide overproduction and cytochrome c release. This mechanistic framework was largely established by the foundational clinical work of Brealey et al5 The transitional burst phase (2005–2017) was characterized by keywords including “multiple organ failure” (2005–2017) and “cytochrome c oxidase” (2005–2015), reflecting a shift toward integrated organ-level investigations. The most clinically significant findings relate to the recent burst phase (2021–2025). “NLRP3 inflammasome” (burst strength 12.21, 2021–2025) has emerged as one of the most active current hotspots. Disruption of mitochondrial quality control creates a feed-forward cycle of NLRP3 overactivation that drives multi-organ inflammation in sepsis.22 “Sepsis-induced cardiomyopathy” and “myocardial injury” (burst strengths 10.87–12.32, 2022–2025) represent dominant organ-specific frontiers. Mitochondrial dysfunction—including disruption of calcium homeostasis, impaired oxidative phosphorylation, and pathological mitochondrial fission—is now recognized as a central contributor to septic cardiomyopathy.23 “Macrophage polarization” (burst strength 8.79, 2023–2025) reflects the growing recognition that mitochondrial metabolism is a master regulator of macrophage functional phenotype in sepsis, with mitochondrial dysfunction exacerbating the imbalance between M1 and M2 polarization states.24 Limitations and Comparison with Related Bibliometric Studies Several limitations should be acknowledged. First, the analysis was restricted to the Web of Science Core Collection, potentially missing publications indexed exclusively in Scopus, PubMed, or Embase. Second, bibliometric analyses are inherently quantitative and do not capture qualitative significance, replication status, or clinical applicability of individual publications. Third, citation counts are subject to citation lag, particularly for publications from 2020 onwards. Fourth, despite careful standardization, ambiguities in institutional affiliations may have introduced minor inaccuracies. Fifth, the thematic map and keyword analyses are based on author-assigned keywords, which may not uniformly reflect the conceptual content of all publications. Sixth, the exclusion of 46 non-English publications, while necessary to ensure consistent bibliometric processing, may have introduced a degree of language bias. Given that China was the leading contributor to publication volume in this field, the omission of Chinese-language literature may have led to a partial underestimation of the true academic output and scholarly influence of Chinese institutions and researchers. This limitation should be considered when interpreting the geographic distribution of research activity, and future analyses incorporating multilingual databases may provide a more complete representation of national contributions. The present study represents the first comprehensive bibliometric analysis of the intersection of sepsis and mitochondrial biology. The bibliometric analysis of sepsis-associated encephalopathy by Zhang et al (Front Med, 2022)10 identified neuroinflammation and blood-brain barrier disruption as dominant themes, consistent with the Niche Theme cluster identified in the present thematic map. The bibliometric analysis of sepsis-associated acute kidney injury by Chen et al (Front Med, 2025)9 similarly identified China and the United States as leading contributors and mitochondria as an emerging research keyword, corroborating the prominence of the mitochondria-AKI Motor Theme. The bibliometric analysis of endothelial cells and sepsis by Shi et al (Heliyon, 2024)12 employed a comparable methodological framework and similarly identified a tripartite co-citation knowledge structure. The present study further differentiates itself through the application of Bradford’s Law journal stratification, Lotka’s Law author productivity analysis, and bibliographic coupling network analysis. Conclusion This comprehensive bibliometric analysis of 3545 publications provides the first systematic overview of global research at the intersection of sepsis and mitochondrial biology over 25 years (2000–2025). The field has experienced sustained and accelerating growth, particularly following the publication of the Sepsis-3 consensus definition in 2016. China and the United States emerged as the dominant contributors in terms of publication volume and citation impact, respectively. Bradford’s Law analysis identified 30 core journals, led by Shock, Frontiers in Immunology, and Critical Care Medicine. Lotka’s Law analysis confirmed that scientific productivity follows the expected inverse square distribution. The co-citation network revealed three foundational knowledge clusters anchored by the seminal mechanistic work of Brealey, Crouser, and Singer, while the bibliographic coupling network identified current research frontiers focused on organ-specific complications including sepsis-associated encephalopathy, sepsis-induced cardiomyopathy, and acute kidney injury. Keyword burst analysis demonstrated a coherent temporal evolution from early investigations of nitric oxide-mediated mitochondrial inhibition, through integrated organ-level studies, to the current focus on NLRP3 inflammasome activation, macrophage polarization, and organ-specific mitochondrial injury. The identification of these emerging topics suggests priority areas for future investigation and highlights opportunities for enhanced international institutional collaboration. Future studies should build upon these findings by incorporating multi-database analyses and integrative approaches combining bibliometric methods with systematic review methodologies.