A literature search was conducted using the Scopus database on 21^st July 2025. The Scopus database was selected because it provides a comprehensive collection of global scientific research, extensive coverage across various research categories, a higher number of publications, and a greater citation count [20,21]. Additionally, it has been widely used in bibliometric analyses. The search strategy employed was as follows: Article title = (“neurotransmitter*” OR “5-HT” OR “serotonin” OR “dopamine” OR “adrenergic” OR “norepinephrine” OR “epinephrine” OR “acetylcholine” OR “ACh” OR “cholinergic” OR “muscarinic” OR “aspartate” OR “aspartic acid” OR “neurokinin*” OR “histamine” OR “monoamine*” OR “oxytocin” OR “glycine” OR “glutamate” or “GABA” OR “gamma-aminobutyric acid” OR “NMDA” OR “N-methyl D-aspartate” OR “AMPA” OR “α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid”) AND (“stress”) AND NOT (“glycine max” OR “glycine soja” OR “monosodium glutamate” OR “glycine betaine”). While restricting the search to article titles enhances specificity and relevance, this approach may have excluded potentially relevant studies in which key terms appear only in abstracts or keywords. This limitation should be taken into account when interpreting the findings. The search was limited to publications from 2000 to 2023. The search date (21st July 2025) served as the data extraction “frozen point” for citation counts. This time frame was selected to ensure the completeness and stability of citation data, as more recent publications may be subject to ongoing indexing and citation updates. All document types except retracted and erratum publications were included in this analysis. Figure 1 demonstrates the screening flowchart of the documents.

Documents were extracted from the Scopus database in Microsoft Excel (.xlsx), Research Information Systems (.ris), and Comma-Separated Values (.csv) formats for analysis. To ensure the accuracy of the results, two independent researchers conducted the literature selection and data extraction. Any discrepancies between the two reviewers were resolved through discussion until consensus was reached. Microsoft Excel was used to calculate document percentages and generate descriptive graphs. Citation metrics were computed using Harzing’s Publish or Perish (PoP) software, while bibliometric mapping and visualization were performed with VOSviewer version 1.6.20 (Universiteit Leiden, Netherlands). VOSviewer provides overlay visualization maps in which node color and distance reflect temporal distribution and relational associations within the research network [22]. The bibliometric indicators analyzed included total publications, total citations, citation trends, subject categories, leading journals, authors, countries, institutions, and keyword co-occurrence patterns.

Of the 6,323 records initially identified and screened, 4,189 fell within the time period from 2000 to 2023. After excluding 54 errata and 3 retracted documents, a total of 4,132 publications were included in the analysis. Original articles accounted for the highest document type (3,674, 88.9%), followed by review articles (264, 6.4%), conference papers (66, 1.6%), book chapters (59, 1.4%), letters (23, 0.6%), and notes (20, 0.5%). Other types of documents, such as short surveys, editorials, books, and data papers, made up less than 0.6% of the total. Among the included publications, 3,959 (95.6%) were in English, while 78 were in Chinese, 36 in Russian, 11 in German, 10 in French, 8 in Portuguese, 7 in Japanese, 5 in Korean, 5 in Persian, 4 each in Spanish and Ukrainian, 2 each in Dutch, Italian, Slovak, and Turkish, and 1 each in Arabic, Czech, Hebrew, Moldavian, Moldovan, Polish, and Romanian.

Figure 2 illustrates the annual publications and citations from 2000 to 2023. Over the past 23 years, publications have steadily increased, reflecting growing interest in the topic and sustained research activity. We anticipate this upward trend will continue. Between 2000 and 2023, the 4,132 publications received 160,485 citations, resulting in an average citation rate of 6,686.88 cites/year and 38.91 cites/paper. The highest annual citation count was in 2014, with 10,815 citations, followed by 2006 with 10,729 citations, and 2012 with 10,711 citations. Of the 4,132 publications, 210 (5.1%) have never been cited, and 701 (16.9%) have been cited fewer than five times. Many of these less-cited publications are relatively recent, published within the past decade.

According to Scopus classification, the 4,132 publications were distributed across 26 subject areas (Table 1). Neuroscience (38.2%), Medicine (38.1%), and Biochemistry/Genetics/Molecular Biology (37.7%) were the most represented categories, followed by Pharmacology/Toxicology/Pharmaceutics (18.6%), Agricultural and Biological Sciences (11.4%), and Psychology (6.4%). Other areas, including Immunology and Microbiology, Chemistry, and Environmental Science, contributed smaller proportions. The other subject categories contributed to less than 9% of the overall publications under the categories of Chemical Engineering, Veterinary, Nursing, Computer Science, Arts and Humanities, Engineering, Social Sciences, Health Professions, Materials Science, Physics and Astronomy, Dentistry, Earth and Planetary Sciences, Energy, Mathematics, Business, Management and Accounting, and Decision Sciences. This broad disciplinary spread reflects the multidisciplinary nature of stress and neurotransmitter research.

The analysis identified the leading contributors across journals, authors, highly cited publications, countries, and institutions (Table 2). The top five journals publishing stress and neurotransmitter research were Neuropsychopharmacology (72 articles), Behavioural Brain Research (71), Journal of Neuroscience (64), PLOS One (64), and Brain Research (62). These journals primarily focus on behavioural neuroscience, psychopharmacology, and neurobiology, aiming to explore the neural mechanisms underlying behaviour, the impact of drugs on both animal and human behavior, and the mechanisms underlying these effects.

Among the 4,132 documents analyzed, 131 were written by a single author, 371 by two authors, 480 by three authors, and 3,150 by more than three authors. The document with the most authors had 47 contributors. Among individual researchers, Popoli M. was the most prolific, contributing 18 publications with 1,890 citations (Table 1). Grace A.A. and Musazzi L. each published 17 articles, followed by Kvetnansky R. (15) and Puglisi-Allegra S. (14). These authors, largely from neuroscience and psychophysiology backgrounds, also demonstrated high citation impact, underscoring their influence in shaping the field. They remain active in research and continue to publish, with nearly all their publications being cited. Notably, there are close collaborations between authors from the same institution, such as Popoli M. and Musazzi L. of Università degli Studi di Milano in Milan.

Several landmark papers have significantly influenced this research area. The most-cited article is “Essential Role of BDNF in the Mesolimbic Dopamine Pathway in Social Defeat Stress” [23], with 1,740 citations, published by Science in 2006 (Table 1). The second most-cited article is “Social Support and Oxytocin Interact to Suppress Cortisol and Subjective Responses to Psychosocial Stress” [24], which received 1,569 citations and was published in Biological Psychiatry in 2003. The third most-cited article, “Pharmacological Inhibition of Cystine-Glutamate Exchange Induces Endoplasmic Reticulum Stress and Ferroptosis” [25], published by eLife in 2014, has 1,368 citations. Of the top ten most-cited publications, only one is a review article, which is “The Stressed Synapse: The Impact of Stress and Glucocorticoids on Glutamate Transmission” [27], while the others are original research articles. The high citation counts of these works likely reflect the originality of their findings and the novel insights they provide, highlighting the central role of neurotransmitter pathways in stress-related mechanisms.

At the national level, the United States dominated output (1,312 publications; 75,301 citations), followed by China (685; 16,322), Japan (279; 9,407), Germany (240; 14,663), and Canada (192; 11,088) (Table 1). Collectively, these five countries accounted for most of the global productivity and citation impact, reflecting both breadth and depth of research activity. The most active institutions were McGill University (52 publications; 4,094 citations), the U.S. National Institutes of Health (49; 3,443), Yale School of Medicine (49; 6,800), Harvard Medical School (45; 2,118), and the University of Pittsburgh (43; 2,561). These findings indicate strong contributions from North American institutions, particularly in neuroscience and psychiatry.

Co-authorship network analysis is a vital tool for examining research collaboration among authors. This study utilized VOSviewer for analysis, selecting co-authorship as the analysis type and authors as the unit of analysis. Parameters were set with a maximum of 25 authors per document and a minimum of 3 documents per author. From a total of 16,415 authors, 381 met these criteria. However, not all these 381 authors are interconnected. Figure 3 (A1 and A2) illustrates the co-authorship network and overlay visualization for the 219 authors. The network visualization map uses colour, circle size, and line thickness to represent clusters, productivity size, and relative link strength (collaboration), respectively [28]. Authors with closer collaborative relationships are depicted with circles that are closer together. Total link strength reflects the cumulative strength of an author’s co-authorship connections with other authors.

The analysis identified ten clusters representing distinct groups of authors in the field of stress and neurotransmitter research. Cluster 1 (red) comprises 28 authors, Cluster 2 (green) also consists of 28 authors, Cluster 3 (blue) includes 27 authors, Cluster 4 (yellow) contains 26 authors, Cluster 5 (purple) has 20 authors, Cluster 6 (aqua) also has 20 authors, Cluster 7 (orange) includes 19 authors, Cluster 8 (brown) comprises 18 authors, Cluster 9 (pink) contains 17 authors. Cluster 10 (grey) includes 16 authors. Notably, certain clusters, such as Cluster 7 and Cluster 4, are distinctly separate from the others, indicating their weak or distant relationship with the remaining clusters. The overlay visualization reveals that Clusters 1, 4, and 8 were among the pioneers in this research domain, as evidenced by their older publications. Conversely, Clusters 3, 6, and 9 were the emerging clusters, with average publication years ranging from 2016 to 2023.

Among the 115 countries that contributed to the published documents, 48 met the threshold, each with at least 10 documents and 10 citations. Figure 3 (B1 and B2) illustrates the co-authorship network and density visualization analysis by country. The network map includes 48 circles, each representing a country. The top five most productive countries—namely, the United States, China, Japan, Germany, and Canada—are depicted with larger circles, indicating higher publication output than other countries. The analysis identified three distinct clusters representing groups of countries in the field of stress and neurotransmitter research. The proximity of the circles indicates closer research collaborations. The largest cluster (red) comprises 21 countries, including Argentina, Austria, Belgium, Brazil, Canada, Chile, Croatia, France, Germany, Greece, Ireland, Israel, Mexico, the Netherlands, Portugal, Serbia, Spain, Sweden, Switzerland, Ukraine, and the United Arab Emirates. The second-largest cluster (green) comprises 16 countries: China, the Czech Republic, Egypt, Hungary, India, Iran, Japan, Pakistan, Poland, Romania, Saudi Arabia, Slovakia, South Korea, Thailand, Turkey, and the United States. The third cluster (blue) includes 11 countries: Australia, Denmark, Hong Kong, Italy, Jordan, Malaysia, Nigeria, the Russian Federation, South Africa, Taiwan, and the United Kingdom. The cluster density visualization highlights denser areas with a high concentration of nodes. The analysis indicates that research on stress and neurotransmitters is most concentrated in the United States, China, and Germany.

Figure 3 (C1 and C2) displays the network and overlay visualization analyses of co-occurrence among author keywords. A minimum threshold of 10 occurrences was applied, revealing that 55 of the 7,436 keywords met the criteria. In addition to stress and neurotransmitters, commonly used keywords included oxytocin, dopamine, serotonin, glutamate, oxidative stress, depression, and anxiety. The analysis identified three distinct clusters within the author’s keyword network. The red cluster (cluster 1, 25 items) contains terms such as acetylcholinesterase, BDNF, behaviour, brain, chronic stress, depression, female, gene expression, genetics, heart rate, hippocampus, human, inflammation, memory, mouse, neurotransmitter, oxytocin, posttraumatic stress disorder, pregnancy, rats, serotonin transporter, sex difference, social stress, stress, and β-adrenergic receptor. The green cluster (cluster 2, 16 items) includes keywords such as acute stress, amygdala, animal model, anxiety, corticosterone, dopamine, hypothalamus, immobilization stress, mental stress, metabolism, microdialysis, norepinephrine, nucleus accumbens, prefrontal cortex, serotonin, and γ-aminobutyric acid. The blue cluster (cluster 3, 8 items) features keywords such as acetylcholine, Alzheimer, antioxidants, apoptosis, glutamate, glutathione, hydrogen peroxide, lipid peroxidation, neuroprotection, neurotoxicity, oxidative stress, Parkinson, reactive oxygen species, and superoxide dismutase. The overlay visualization analysis indicates that newer keywords (from 2016 onwards) include antioxidants, neuroprotection, inflammation, BDNF, γ-aminobutyric acid, acetylcholinesterase, metabolism, and gene expression.