DOI: https://doi.org/10.21203/rs.3.rs-2278919/v1
Post-stroke depression(PSD) is one of the most common complications of stroke which affects recovery and quality of patients’ life adversely. PSD also is a significant source of burden for caregivers. Recently, the number of publications on PSD has been an increasing worldwide that means PSD has attracted more and more attention. Therefore, a more comprehensive understanding of the publications on PSD is now required. The purpose of this paper is to analyze the research status, discover the hotspots and frontiers about PSD.
The publications’ raw data was retrieved from the Web of Science Core Collection on September 19, 2022. Impact Factor (IF) and Journal Citation Reports (JCR) segments of the publications were obtained from the Web of Science website. Visualization analysis was performed on CiteSpace and VOSviewer.
From 2012 to 2022, the number of citations and publications about PSD increased exponentially year by year. Finally, a total of 3268 publications related to PSD were identified. China (867) and the US (738) had the most publications; Capital Med Univ and Univ Melbourne were the most prolific institutions. Ungvari GS and Hackett ML were the most active authors and the most prolific and co-cited, respectively. Journal of stroke cerebrovascular diseases (131 articles) was the highest output. Among the 3268 publications, International journal of stroke had the most citations (43.9 times), Stroke has the most citations (2,990) and the highest H-index value. The citation burst for 8 co-cited references lasts until 2022. Finally, this paper divided the hot spots into four categories: cluster 1 (study on risk factors of PSD), cluster 2 (clinically relevant research on PSD), cluster 3 (mechanisms of PSD), and cluster 4 (the Treatment Study for PSD).
PSD research is one research main area worldwide. This work demonstrated the knowledge structure and the evolution of the PSD research field and predictdresearches trends through visual analysis. The study of clinical treatment and prevention of PSD will instead of the study of the mechanism in the future. This study provides reference for future research of PSD.
Post-stroke depression (PSD) is a common, clinically significant neuropsychiatric symptom because it adversely affects recovery and quality of stroke patients’ life, interferes with recovery, and is a significant source of burden on caregivers(Robinson and Jorge, 2016).The term "post-stroke depression" is often used to define any depressive state that occurs after a stroke, regardless of the time or symptoms(Das and G, 2018). Empirical evidence of pathophysiological factors associated with PSD suggested that it was closely associated with excess proinflammatory cytokines(Chen et al., 2020; Shu et al., 2022), hypothalamic-pituitary-adrenal (HPA) axis dysfunction, and altered neuroplasticity, even this pathogenesis has been speculative,that is whyeffective treatments for PSD remain limited (Medeiros et al., 2020).Recently, more and more researchers are studying PSD to give the high incidence of PSD.
However, there is no quantitative analysis of publications on PSD. In the face of such a large number of publications, it is important to summarize, analyze and understand the main research results in this field, and determine the research focus and future trend in this field. Bibliometric analysis can help readers better understand research hotspots and future research directions in this field by visualizing the content and citations of publications, and provides important references for subsequent studies. However, there are no complete bibliometric studies in the field of post-stroke depression. Bibliometrics provides an intuitive way to understand the overall structure and evolution trend of a research field, which is agreat significance for us to grasp the knowledge framework and research hotspots in the field of PSD (de Castilhos et al., 2020; Liu et al., 2022).
This study is based on a bibliometric analysis of publications on PSD from 2012 to 2022 using the Web of Science database, CiteSpace and VOS viewer software (Wang et al., 2022).Specifically, the main objective of this paper was to summarize the current research status and major contributors of PSD (including analysis of annual publications, countries, institutions, journals, authors, keywords, keyword clustering, keyword emergence, etc.), this study summarized the main research hotspots in this field, and look into the future of this field, providing a brand new approach to understand this field from a professional perspective, so that readers can better grasp the research status and research trends in this field (Ma et al., 2021).
The Web of Science (WoS) Core Collection database was used to perform the bibliometric literature search from January 1, 2012 to September 09, 2022. The search formula was set as follows:"Posts-troke Depression(Topic)" or "Poststroke Depression(Topic)" or "Depression after cerebral infarction(Topic)" or "Ischemic Stroke Depression(Topic)".
Finally, 3268 records (2694 articles and 574 reviews) were analyzed and summarized. English articles and review publications related to PSD were included, other document types were excluded (Shen et al., 2022). The recording flow chart was shown in Fig. 1.
CiteSpace is a visualization software for bibliometric analysis which was developed by Professor Chen Chaomei from Drexel University. CiteSpace 6.1. R3 was used to analyze the final record in this study. The parameters included: time slice (2012–2022), years per slice (1 year), selection criteria (g-index, k = 25) and pruning (pathfinder and pruning the merged network). Other parameters of different situations were set according to the CiteSpace manual. VOS viewer software, a useful tool for building and visualizing bibliometric networks was developed by the Science Center at Leiden University (Netherlands) in 2007.The latest version 1.6.18 was available for free download (https://www.vosviewer.com/). In the VOS viewer software, different parameters, such as country, journal, institution, keyword, etc. were represented by each node, and the weight of the parameters, such as the number of publications, citations, or frequency of occurrence, would determine the size of the node. A higher weight will get a larger node. Nodes and lines were colored by the cluster they belong to. Lines between nodes represent links. Link strength was assessed by the Total Link Strength Index (TLS), which was the sum of all link strengths and can be extended to reflect link strength between institutions. Additional information such as Journal Impact Factor (IF) and Journal Citation Report (JCR) were available directly from the Web of Science website on September 19, 2022. Analyze annual publications with Microsoft Office Excel 2019.
Cluster view and burst detection: Execute the cluster view on the generated graph, and annotate the cluster view by referring to the citing title, keywords and subject. The function of burst detection was to detect the large change in the number of citations within a certain period of time, which can reflect the decline or rise of keywords.
Double Graph Overlay: Double Graph Overlay is a new way to show the distribution and citation trajectories of articles across disciplines. As a result, there is a distribution of citing journals on the left and cited journals on the right. Curves were citation lines that fully show the context of the citation.
As can be seen from Fig. 2, the annual publication volume and citation volume of the WoSCC database show an overall upward trend from 2012 to 2022. Prior to 2016, the research on post-stroke depression was relatively slow, with no more than 250 papers published each year and no more than 3,000 papers cited each year. However, the number of annual publications and citations had gradually increased after 2016.
In 2021, the annual publication volume reached 461 and the number of citations reached a peak of 12,351,means that there were more and more relevant studies in this field.
The top 10 countries/regions in the WoSCC database by the number of articles on PSD were shown in Table 1. The top three countries in the field of PSD were China (26.52%), the United States (22.58%) and Australia (8.26%), accounting for about 57.36% of the total. The United States was the country with the highest Total citations and H-index, while the country with the highest Average citations per article is Italy.
Rank | Countries/Regions | Article counts | Percentage (n/3268) | Total citations | Average citations per article | H-index | TLS |
---|---|---|---|---|---|---|---|
1 | China | 867 | 26.52% | 8926 | 10.3 | 37 | 256 |
2 | USA | 738 | 22.58% | 16438 | 22.21 | 58 | 509 |
3 | Australia | 270 | 8.26% | 6220 | 23.04 | 38 | 350 |
4 | England | 245 | 7.49% | 7554 | 30.83 | 43 | 305 |
5 | Germany | 214 | 6.54% | 6498 | 30.36 | 41 | 259 |
6 | South Korea | 172 | 5.26% | 2635 | 15.32 | 27 | 115 |
7 | Canada | 172 | 5.26% | 4328 | 25.16 | 37 | 166 |
8 | Netherlnds | 150 | 4.58% | 3739 | 24.93 | 32 | 164 |
9 | Italy | 132 | 4.03% | 4638 | 35.14 | 32 | 174 |
10 | Japan | 118 | 3.61% | 1527 | 12.94 | 21 | 89 |
Figure 3A was a map of international cooperation between countries/regions, the thicker the lines between two countries indicate the closer cooperation. As the graph shown, countries such as China, the United States and Australia were more closely connected with other countries. Figure 3B was the country's citation network visualization map. Countries with total link strength (TLS) over 300 were USA (TLS = 509), Australia (TLS = 350) and England (TLS = 305), indicated that these three countries were more influential internationally. Overall, China, USA and Australia were the major international contributors to PSD research, with the largest number of publications and higher quality papers. Although the number of publications in China was relatively high, the quality of publications, as well as research in this field, needs to be improved.
The top 10 institutions with the publications’ number were illustrated in Table 2. China, Australia and Canada had the most active institutions in the field of PSD research. Capital Medical University, University of Melbourne, University of Toronto and University of Western Australia were the top four institutions with the largest number of published papers, and University of Toronto had the highest number of citations and H-index.
Figure 4(A) was an institutional network collaboration diagram created using CiteSpace. Each In this graph, different institutions were represented by different nodes, with darker nodes indicating later active years. Capital Medical University and Harvard University had the highest centrality (both 0.13), followed were University of Toronto and Chinese University of Hong Kong (both 0.12). When the centrality value is greater than or equal to 0.1 means that the node is a key node of the network graph. That means these four universities have an important position in institutional cooperation. The institution's citation network visualization map was created by VOS viewer, as shown in Fig. 4(B), with 2699 links and 402 nodes, forming 14 clusters of different colors. University of Western Australia (TLS = 182), University of Melbourne (TLS = 180) and University of Toronto (TLS = 126) were the top three institutions with the highest TLS.
Rank | Institutions | Countries/Regions | Article counts | Total citations | Average citations per article | H-index |
---|---|---|---|---|---|---|
1 | Capital Medical University | China | 59 | 787 | 13.34 | 15 |
2 | University of Melbourne | Australia | 58 | 1037 | 17.28 | 18 |
3 | University of Toronto | Canada | 57 | 1684 | 26.73 | 23 |
4 | University of Western Australia | Australia | 51 | 1173 | 20.58 | 18 |
5 | Chinese University of Hong Kong | China | 51 | 832 | 16.31 | 17 |
6 | Wenzhou Medical University | China | 51 | 556 | 10.69 | 15 |
7 | Maastricht Univ | Netherlands | 41 | 764 | 18.19 | 14 |
8 | Kings Coll London | England | 40 | 2531 | 55.02 | 23 |
9 | Harvard Med Sch | USA | 40 | 798 | 19.95 | 14 |
10 | Harvard Univ | USA | 34 | 2813 | 29.93 | 28 |
The top 10 funding agencies which were supporting researches on PSD were listed in Table 3, and United States Department of Health Human Services (316), the National Institutes Of Health (306) and the National Natural Science Foundation Of China (300) were the top three. The top three funding agencies for post-stroke depression disease support far more than any other agency. By country, the United States China and the European Union funded the most publications.
Rank | Funding Agencies | Number of Publications | Countries/Regions |
---|---|---|---|
1 | United States Department Of Health Human Services | 316 | USA |
2 | National Natural Science Foundation Of China | 300 | China |
3 | European Commission | 131 | European |
4 | National Institute Of Neurological Disorders Stroke | 114 | USA |
5 | National Health And Medical Research Council Nhmrc Of Australia | 63 | Australia |
6 | Canadian Institutes Of Health Research | 51 | Canada |
7 | German Research Foundation | 51 | Germany |
8 | Medical Research Council | 46 | England |
9 | Uk Research Innovation | 46 | England |
10 | National Research Foundation Of Korea | 46 | South Korea |
The top 10 authors published a total of 299 articles on post-stroke depression, accounting for 9.14% of the total publication volume, see Table 4. Five of the top 10 authors were from China, namely Wang J, He JC, Tang WK, Zhang Y, Liu Y;, three from Korean namelyCho KH, Kim JS, and Kim JM. Among these authors, Ungvari GS and Wang J co-authored three articles, namely "Association between schizophrenia and violence among Chinese female offenders" and "Nurses' work-related stress in China: a comparison" between psychiatric and general hospitals" and "The MacArthur Competence Assessment Tools for assessing decision-making capacity in schizophrenia: A meta-analysis".
Figure 5(A) was a co-citation network graph of authors who had been cited at least 20 times, with a total of 957 nodes, 148,077 links and 5 clusters. Robinson RG (TLS = 30719), Hackett ML (TLS = 30477) and Dreier JP (TLS = 21729) had the highest TLS. The co-author citation analysis visualization graph, shown in Fig. 5(B), contains a total of 10,879 nodes, 10,879 links and 8 clusters. The nodes in this graph represent authors, the difference between the nodes in Fig. 5(A) and Fig. 5(B) was that in Fig. 5(B), the collaboration between authors determined the links between nodes. Hackett ML (TLS = 1439), Dreier JP (TLS = 1222) and Ungvari GS (TLS = 745)were the top 3 authors with the highest TLS, who were at the center of the partnership. Overall, the nodes of the network graph were scattered, indicating that the collaboration between authors in this field was not close.
Rank | Author | Count | Countries/Regions | Institutions | H-index | Toyal citations |
---|---|---|---|---|---|---|
1 | Ungvari GS | 37 | Australia | Univ Notre Dame Australia | 15 | 638 |
2 | Wang J | 33 | China | Affiliated Hosp Youjiang Med Univ Nationalities | 12 | 537 |
3 | He JC | 32 | China | Wenzhou Med Univ | 11 | 325 |
4 | Tang WK | 32 | China | Chinese Univ Hong Kong | 13 | 474 |
5 | Zhang Y | 32 | China | Shanghai Jiao Tong Univ | 10 | 276 |
6 | Cho KH | 28 | South Korea | Kyung Hee Univ | 15 | 628 |
7 | Dreier JP | 27 | Germany | Charite Univ Med Berlin | 18 | 1225 |
8 | Liu Y | 27 | China | Shanghai Univ Tradit Chinese Med | 7 | 134 |
9 | Kim JS | 26 | South Korea | Univ Ulsan | 11 | 427 |
10 | Kim JM | 25 | South Korea | Chonnam Natl Univ | 14 | 544 |
The top 10 journals by publication volume from 2012 to 2022 were listed in Table 5, and most of them from the United States. The journal with the most publications was Journal of stroke cerebrovascular diseases, followed were Stroke and Topics in stroke rehabilitation, Stroke had the highest impact factor (IF). The IF of the top 10 journals range from 1.897 for Topics in stroke rehabilitation to 7.19 for Stroke according to Journal Citation Reports (JCR) 2022. The visualization of co-citation analysis of journals was shown in Fig. 6(A). The top three journals with the highest TLS were Stroke (TLS = 853), Journal of stroke cerebrovascular diseases (TLS = 814) and International journal of stroke (TLS = 805). Figure 6(B) show a double-map overlay of all academic journals, the right and left side of the map represented the cited and citing journals respectively, the colored lines represented the citation relationship between the citing journals and the cited journals. The entire graph can show the complete citation process. The length of the vertical axis was determined by the number of papers published in the journal, and the length of the horizontal axis was determined by the number of authors.The statistical results shown that Medicine, Medical and Clinical were the three major fields that published articles focus on, while the cited journals were mainly published in the fields of Psychology, Education and Social.
Rank | Journal title | Countries/Regions | Article Counts | Percentage(N/3268) | IF(2022) | Quartile in category | H-index |
---|---|---|---|---|---|---|---|
1 | Journal of stroke cerebrovascular diseases | USA | 131 | 4.00% | 1.787 | Q4 | 22 |
2 | Stroke | USA | 100 | 3.05% | 7.19 | Q2 | 33 |
3 | Topics in stroke rehabilitation | USA | 84 | 2.57% | 1.897 | Q3 | 19 |
4 | Plos one | USA | 68 | 2.08% | 2.74 | Q2 | 25 |
5 | Medicine | USA | 64 | 1.95% | 1.552 | Q4 | 10 |
6 | Disability and rehabilitation | England | 63 | 1.92% | 2.222 | Q2 | 19 |
7 | Journal of affective disorders | Netherland | 60 | 1.83% | 3.892 | Q3 | 19 |
8 | Frontiers in neurology | Switzerland | 59 | 1.80% | 2.889 | Q3 | 13 |
9 | Neuropsychiatric disease and treatment | New Zealand | 46 | 1.40% | 2.157 | Q4 | 10 |
10 | Internationaljournal of stroke | England | 43 | 1.31% | 4.882 | Q3 | 20 |
A visualization of the literature co-citation network generated by the VOS viewer in Fig. 7(A), and the top 10 most-cited articles of PSD research were listed in Table 6, with 65,349 links and 587 nodes, forming 6 clusters of different colors. In these clusters, the total link strength to other cited references was calculated, and the highest TLS was an article published in 2005 by Hackett, ML et al. (TLS = 4734), followed ones were Ayerbe L et al. (TLS = 4734)., Hackett ML et al. (TLS = 4645, 2014), and Robinson RG et al. (TLS = 3955, 2016),. Two of the 10 articles were published on Stroke, and four of them were published between 2013 to 2017. The most cited article of PSD was published by Hackett, ML et al. about understanding the predictors of stroke-related depression that may lead to better treatment, including prevention and treatment.
The top 20 references with the strongest citing outbreaks in the PSD study were shown in Fig. 7(B). Among them, the two most explosive articles with strengths over 30 were: Natural history, predictors and outcomes of depression after stroke: Systematic review and meta-analysis which was published by Ayerbe, L et al. in 2013;and Post-stroke depression: Mechanisms and pharmacological treatment, which was published by Villa, RF et al. in 2018. The primary predictors of depression were identified in the first article, requiring intervention for PSD and its underlying outcomes. The second article had a bidirectional association between depression and stroke, and related research on the mechanisms and drug treatment of PSD. Both articles contributed significantly to post-stroke depression, making them with the strongest outbreaks. Robinson RG et al. and Hackett ML et al. each published an article that broke out in 2008, and the most recent outbreak occurred in 2016, and it continued to today.
Rank | Title | Total citations | First author | Publication Year | Journal |
---|---|---|---|---|---|
1 | Frequency of depression after stroke: an updated systematic review and meta-analysis of observational studies | 4645 | Hackett, ML | 2014 | International Journal Of Stroke |
2 | Post-Stroke Depression:A Review | 3955 | Robinson,RG | 2016 | Cerebrovascular Diseases |
3 | Natural history,predictors and outcomes of depression after stroke:systematic review and meta-analysis | 4669 | Ayerbe,L | 2013 | British Journal Of Psychiatry |
4 | Predictors of depression after stroke-A systematic review of observational studies | 4734 | Hackett,ML | 2005 | Stroke |
5 | The hospital anxiety and depression scale | 2913 | Zigmond,A S | 1983 | Health And Quality Of Life Outcomes |
6 | Measurements of acute cerebral infarction:a clinical examination scale | 2305 | Brott,T | 1989 | Stroke |
7 | "Mini-mental state".A practical method for grading the cognitive state of patients for the clinician | 2114 | Folstein,M F | 1975 | Journal Of Psychiatric Research |
8 | The hospital anxiety and depression scale | 2219 | Zigmond,A S | 1965 | Health And Quality Of Life Outcomes |
9 | A rating scale for depression | 2186 | Hamilton,m | 1960 | Journal Of Neurology Neurosurgery And Psychiatry |
10 | The hospital anxiety and depression scale | 2272 | Zigmond,A S | 2017 | Acta Psychiatrica Scandinavica |
The evolution of keywords in time series was an important basis for the development of this field in different time periods, and could predict the future research direction (Li et al., 2022). Figure 8 shown the bubble chart of keywords of high frequency topics from 2012 to 2022. By sorting and combining keywords and similar keywords, while retaining the most representative topics in the field of PSD research, high-frequency keywords in the field of PSD research were finally summarized into 12 hot topics.The size and color of the bubbles clearly indicate the evolution of keywords over the years. Therefore, the topics can be divided into three categories: the first category includes: "quality of life", "risk factor", "symptom", "predictor" and "risk", related to the first category of hot topic keywords. There were more and more articlesdepending on the status quo shown by the icon, these topics may receive more attention in the future. The second category included "prevalence", "scale" and "meta analysis". The frequency of occurrence first maintained an upward trend, and entered a stable period after 2020. The frequency fluctuated slightly in recent years, indicating that it had still received corresponding attention in recent years. The third category was divided into "cognitive impairment", "recovery", "disorder", and "rehabilitation". In the past 4 to 5 years, there had been a downward trend, indicating that the degree of attention to this topic haddecreased significantly.
In bibliometrics, keywords are a very important part, and keyword co-occurrence analysis can reveal changing research topics and development trends(Wang et al., 2022). The density visualizations generated by keyword co-occurrence more than 10 times were shown in Fig. 9(A). The higher the frequency of the keyword, the darker the color. 385 keywords were appeared as nodes, with the exception of "brush strokes" and "depression" being included in the search terms. The most frequently occurring keywords were "post stroke depression", "depression", "ischemic-stroke" and "predictors". The overlay visualization map of keywords was shown in Fig. 9 (B), where the keywords that appear more than 30 times at the same time are represented by 135 nodes. In the VOS viewer, different colors are used to label keywords based on the average year (AAY) of the keyword. Blue indicates the keyword that appeared earlier in the time frame, green indicates the second keyword, and yellow indicates the most recent keyword.
The keyword co-occurrence graph was performed basis on Cluster analysis. The more nodes the cluster contains, the smaller the cluster number. Module value (Q value) and average contour value (S value) were used as evidence for judging the effect of graph clustering (Wu et al., 2022).Generally, Q value > 0.3 means the clustering structure was significant; S value > 0.7means the clustering result is convincing. LLR model was used in this study to perform cluster analysis on keywords, as shown in Fig. 10, Q value = 0.418, S value = 0.7317, and the clustering results could be considered significantly .
This study listed the top 20 emergent keywords, as shown in Fig. 11. The keyword with the highest emergent intensity was “trial” (intensity = 9.61), followed were “follow up” (intensity = 7.7) and “community” (intensity = 6.12). "follow up", "infarction", "vascular depression" and "mood disorder" were the earliest emergent keywords, and the emergent keywords from 2020 to 2022 were "guideline", "cohort study", "update" and " statement". The graph show that a total of 8 keywords were still in the ongoing emergent phase.
The bibliometric analysis of PSD over the past 10 years was performed in this study. The number of citations and publications shown a continuous but erratic growth trend year by year. The numbers of published articles (461) and citations (12,351) were the highest in 2021. These results suggested that research involving PSD was gaining worldwide attention.
In terms of Country/Region, China dominated the number of articles published (867), the U.S. (738) and Australia (270) were following China, while the U.S. was far ahead in citations (16,438) Countries, it could be said that China and the United States were in an advantageous position in this field. 4 European countries, 3 Asia-Pacific countries, 2 American countries and 1 Australian country were the top 10 countries with most publications of PSD. 3 Chinese institutions, 2 American institutions, 2 Australian institutions, 1 Dutch institution, 1 British institution and 1 Canadian institution were the top 10 institutions with publications of PSD. According to international cooperation, the University of Western Australia in the United States had relatively close cooperation with other institutions. Although extensive collaborations had been established between countries and institutions, future research involving PSD should focus on international collaborations with multicenter, large-sample studies.
Scientific research and innovation require a large amount of financial, human and material support, which can motivate high-level institutions to make contributions to scientific research. The United States Department of Health Human Services provided financial support for most research projects of PSD, which was one of the reasons why the US has a high academic status in this field.
These prolific German authors have the highest H-index according to the survey of author information, and 5 of the top 10 authors were from China, but the H-index was relatively low, which indicated that their papers’ quality needed to be improved. In terms of authoritative journals, Journal of stroke cerebrovascular diseases (131 papers), Stroke (100 papers), and Topics in stroke rehabilitation (84 papers) contributed the most to the number of published papers. Among the top 10 journals, Q2 accounted for 30%, Q3 accounted for 40%, and Q4 accounted for 30%, indicating that the research quality of PSD still needs to be strengthened. International journal of stroke had the highest average citations (43.9 times). Stroke had the most citations (2,990). The highest H-index value occurd in Stroke. In addition, the IF of the top 10 journals did not exceed 8, and the IF values of 7 journals were 1–3 (Journal of Stroke Cerebrovascular Diseases, Topics in stroke rehabilitation, Plos one, Medicine, Disability and rehabilitation, Frontiers in neurology, Neuropsychiatric disease and treatment), two journals had an IF value of 3–5 (Journal of affective disorders, International journal of stroke), and one journal had an IF value of 7–8 (Stroke). These results suggested that in the future, we should not only pursue quantity, but conduct high-quality research. Analysis of the reference literature shows that the majority of high-quality articles were published on Stroke. As the journal with the highest co-citation rate in the field, Stroke had exerted a significant influence on the academic community of PSD, and its publications serve as authoritative references.
The time evolution of keywords shows that "quality of life", "risk factor", "symptom", "predictor" and "risk". Moreover, through keyword clustering and highlighting, it is not difficult to find that the field of PSD is becoming more and more more attention. Therefore, the academic researches related to these directions in the field of PSD should be further expanded. Over time, based on the trend of the hot keywords and topics, the clinical treatment and prevention research were the new hot spots of PSD instead of early research on the pathogenesis and clinical manifestations.
Clinical Neurology (1014), Neurosciences (452) and Psychiatry (384) were the top three prolific area of research according to the subject category of articles on PSD. The top 10 subject categories were Clinical Neurology, Neurosciences, Psychiatry, Rehabilitation, Peripheral Vascular Disease, Medicine GeneralInternal, Pharmacology Pharmacy, Geriatrics Gerontology, Medicine Research Experimental, Multidisciplinary Sciences indicated that PSD was a complex issue requiring multiple disciplines intervention.
The study of frequently occurring keywords in bibliometrics, may reveal patterns of change and major themes that were critical to understanding the development of the field (Miao et al., 2022). After all the keywords involved were aggregated and analyzed, the data were imported into CiteSpace again and performed keyword clustering analysis, finally got the result of silhouette S = 0.7317 and modularity Q = 0.418. Four clusters were identified in keyword clustering and keyword highlighting (Wang et al., 2021).Then, after reviewing and discussing the relevant literature, the final four cluster names were derived, which were: Cluster 1 (study of risk factors of PSD) was the largest cluster, contains 36 keywords, mainly related to terms such as "risk factors", "quality of life", "outcome assessment", etc.;Cluster 2 (clinical related research on PSD) contained 32 keywords, mainly related to terms such as "cognitive impairment", "poststroke dysphagia", "speech therapy", etc.; Cluster 3 (mechanism study of PSD) contained 25 keywords, including: "oxidative stress", "inflammation", "endothelial dysfunction", etc.; Cluster 4 (Treatment research for PSD) contained 34 keywords, focusing on "transcranial magnetic stimulation", "noninvasive brain stimulation", "robotic training", etc.
Group 1: Study of risk factors for depression after stroke
Study of risk factors for depression after stroke was the largest of the four clusters. Several risk factors have been shown to be strongly associated with the incidence and adverse outcomes of PSD in the research of PSD field. Through decades of extensive research, researchers have discovered that PSD is a multifactorial disease(Babkair, 2017). These factors relate to genetic factors, age, gender, history of depression, severity of stroke, location of the lesion, and in addition to the above risk factors, social support, marriage, and years of education were also associated with PSD which would increase the risk of PSD (Medeiros et al., 2020). Although there are many studies on risk factors for PSD, the factors involved vary from study to study and results vary, so it is not currently possible to standardize risk factors for PSD (Kawada, 2018).
In recent years, various post-stroke depression risk assessment scales have been proposed, such as: Hamilton Depression Rating Scale (HDRS), Patient Health Questionnaire-9 (PHQ-9), Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), Hospital Anxiety and Depression Scale (HADS, scores ≥ 7 means depression) and Center for Epidemiological Studies Depression Scale (CES-D), etc. Among them, HDRS and PHQ-9 were found to be the best to accurately predict the risk of PSD (Trotter et al., 2019). It may aid in the early clinical selection of interventions to manage the risk of PSD (Sarkar et al., 2021). By exploring the risk factors of PSD can help clinicians to improve their understanding of the occurrence of PSD, predict the pathogenic factors more accurately, and then choose prevention and treatment in a timely and effective manner(Wang et al., 2020). This is of great significance to improve the prognosis, shorten the length of hospital stay and reduce the economic burden of PSD patients.
Group 2: Clinically relevant studies of PSD
Many clinical studies are underway to overcome PSD. Curcumin has been shown to inhibit inflammation mediated by CA2 channel activity and reduce the progression of PSD in rats (Camargos et al., 2020).Minocycline had shown antidepressant effects in stroke-induced mice and had been observed to be significant in behavioral improvement (Wang et al., 2019).Therefore, this may be an effective treatment for PSD in the future. The preclinical study of Xiaoyao Jieyu San was a traditional Chinese medicine consisting of 11 components. By increasing 5-HT in PSD rats, the expression of norepinephrine and Brain-Derived Neurotrophic Factor (BDNF) werereduced, thereby reducing the degree of depression in PSD (Trusova and Levin, 2019).Spikes in glutamate levels in PSD were common, and giving pyruvate to experimental rats had been reported to limit the amount of glutamate to reduce symptoms of anxiety and depression (Pei et al., 2019).Studies had shown that aryl hydrocarbon receptor nuclear translocation protein 2 (ARNT2) could help neurons survive and have neural functions. The effect of ARNT2 in PSD rats may be a therapeutic target for further translational research (Hu et al., 2019; Rice et al., 2022). Recent studies by researchers have shown (Abdoulaye et al., 2021)that paeoniflorin (terpenoid glycoside) and fluoxetine can improve depressive symptoms in PSD rats and exhibit high expression of BDNF in the CA1 region of the hippocampus.Ca2/ CAM-dependent protein Kinase II (CaMKII) was increased in PSD rats, and K9M3 inhibitors could inhibit CaMKII expression, thereby reducing PSD symptoms. As an inhibitor of CaMKII, K9M3 may be a potential drug for the treatment of PSD in clinic (Vahid-Ansari and Albert, 2018.Anthocyanin-enriched horse kwai berry extract attenuates depression in mice and mouse models when evaluated in the forced swim test and the tail suspension test. Currently, the lack of appropriate preclinical models mimicking PSD pathology makes research a daunting task. In addition to this, several other limitations and adverse effects limit the use of various drugs (Di Lorenzo et al., 2019).
Group 3: Mechanistic studies of post-stroke depression
The major neuroendocrine stress response system involved in the regulation of mood, immunity, and metabolism because ofdysfunction of the hypothalamic-pituitary-adrenal (HPA) axis(Zhou et al., 2022). The hyperactivity of the HPA axis is one of the most consistent and significant biological findings in major depressive psychiatry (Villa et al., 2018). First, when the hypothalamus receives a signal from the hippocampus or other tissues, the paraventricular nucleus of the hypothalamus released corticotropin-releasing hormone (CRH), which in turn stimulates the pituitary gland to release adrenocorticotropic hormone (ACTH). ACTH affected the synthesis and release of glucocorticoids by the adrenal cortex. Glucocorticoids and their receptors were important components of the HPA axis(Wang et al., 2022). In addition, inflammation and trauma were on the rise(Weina et al., 2018). Inflammation caused by pro-inflammatory cytokines such as interleukin 1α (IL-1α), interleukin 1β (IL-1β), interleukin 6 (IL-6), and tumor necrosis factor-α (TNF-α) is strongly associated with depression (Feng et al., 2022; Xie et al., 2022).Severe depression, elevated serum or plasma levels of IL-6 and C-reactive protein (CRP) were frequently observed by investigators in their studies. Therefore, anti-cytokine therapy was a potential therapy for PSD(Kappelmann et al., 2021; Pericaud et al., 2022).
Currently, there are two theories of neurotransmitter transfer related to PSD, the glutamate-mediated excitotoxicity and the monoaminergic hypothesis. Monoamines, mainly including serotonin (5-HT, or serotonin), dopamine (DA), and norepinephrine (NE), had crucial effects on the brain(Desai et al., 2022; Stanfill et al., 2016; Zhou et al., 2020). Furthermore, the researchers found that 5-HT, NE and DA neurons were directly or indirectly connected in the CNS. However, monoamine levels decreased in people with depression(Casement et al., 2016).Extensive research had justified this hypothesis and was widely accepted. A variety of neurotrophic factors protected neurons from damage caused by depression, which leaded to neuronal atrophy in the limbic region and cerebral cortex(Guo et al., 2022). Insulin-like growth factor-1 (IGF-1) and brain-derived neurotrophic factor (BDNF) were the most important neurotrophic factors, because they affected recovery and neurological rehabilitation after stroke(Luo et al., 2019a).Despite solid evidence for a strong link between depression and stroke(Zhang and Liao, 2020), the explanation for the underlying mechanism remains speculative and whether the link was direct or indirect remains to be determined(Luo et al., 2019b).
Group 4: Treatment Research for Post-Stroke Depression
The study found that nortriptyline improved depressive symptoms better than a placebo(Pluijms et al., 2022). In addition, we found that Tricyclic antidepressants (TCA), such as imipramine and desipramine and mianserin, were significantly improved in patients with PSD(Legg et al., 2020). In the context of PSD treatment, selective serotonin reuptake inhibitors (SSRIs) had been extensively studied and found to be the most safe and effective antidepressants (Mortensen and Andersen, 2021; Sun et al., 2017). However, various adverse effects such as bleeding, cardiovascular side effects, intracranial hemorrhage and metabolic enzyme inhibition limited their use (Trajkova et al., 2019). Fluoxetine was effective against PSD, but its relationship with TCA was unclear. In addition, citalopram and reboxetine were found to be safe for PSD in a randomized trial. However, using different drug treatments for PSD would have conflicting results. Studies had shown that paroxetine and citalopram remain the safest drugs because of their minimal bleeding risk and inhibition of metabolic enzymes (CYP450)(Almuwaqqat et al., 2019; Mortensen and Andersen, 2015).Currently, there were no guidelines for the selection and duration of optimal antidepressant therapy in patients with PSD, and no specific recommendations were provided by the American Stroke Rehabilitation Guidelines (Li and Zhang, 2020).However, according to clinical experience, research and Italian Stroke Prevention and Awareness Diffusion (SPREAD) guidelines (Chertcoff et al., 2021),4–6 months were recommended to continue taking antidepressants, and then gradually stop them. Antidepressants were approved by the American Heart Association (AHA) and at least continued 6 months after recovery (Towfighi et al., 2017). Until now, there is no good medication or psychotherapy for preventing PSD immediately after stroke, early prediction and intervention may have been better for preventing PSD. In recent years, a variety of new treatments for post-stroke depression had emerged. The traditional Chinese medicine Sihogayonggolmoryeo-tang (SGYMT)which have effect on mental disorder, now proving to be a promising alternative to drug therapy for PSD (Kwon et al., 2019). Psychotherapy which includes problem-solving therapy (PST), cognitive behavioral therapy (CBT) and ecosystem focused therapy (EFT) was also selected as an intervention to reduce psychosocial distress in PSD (Hadidi et al., 2017; Kampling et al., 2021; Wang et al., 2018). Problem-solving methods included developing adaptive problem-solving skills and overcoming problems in daily life, aiming to reduce psychological distress and improve the quality of social life. Studies have found that seizures would regulate the nerves of the depressed brain, which was effective for PSD. Electroconvulsive therapy (ECT) can induce seizures through the scalp to pray for a certain therapeutic effect on PSD. Transcranial direct current stimulation (tDCS) involved the use of shallow electrical currents to stimulate specific brain regions that had been found to be very helpful in depression (Jellinger, 2022). Neuromodulation via repetitive transcranial magnetic stimulation (rTMS) involved the non-invasive repetitive generation of magnetic pulses followed by electric fields that induce specific brain regions.When low-frequency rTMS (1HZ) was applied to the left anterolateral prefrontal cortex, the study found improvements in patients' mood and cognitive function, and the observed side effects were smaller than those observed with ECT (Valiengo et al., 2022). Electroacupuncture (EA) was a new treatment for psychiatric disorders that involves electrical stimulation of specific areas through acupuncture to treat a variety of ailments(Cai et al., 2022).Acupuncture, as a TCM intervention, can significantly improved the health and psychological symptoms of patients. miRNA was also used in the prevention and treatment of PSD. The researchers found that rats treated with miRNA mir363-3p showed fewer depressive symptoms and movement disorders compared to the control group(Lv et al., 2022). In conclusion, neuromodulation therapy will be a promising treatment for PSD if a large number of translational safety and efficacy studies can be conducted. Now PSD still faces many challenges. Future studies are needed to investigate whether antidepressants and different combinations of antidepressants can help treat PSD (Zhang et al., 2017). In addition, new developments may point to anti-cytokine modulators as possible treatments for PSD.
Because this work only queried the Web of Science Core Collection database, high quality papers published in journals outside the Web of Science Core Collection database may be missed. In addition, this study only obtained objective results in related research fields through bibliometric analysis, but some potential causes of these results were not further explored. So there are some limitations in this study. The discussion and analysis in this paper was designed to help readers build a clearer framework and implement more effective strategies to better understand PSD. This paper can also help the research community identify existing research gaps, thus contributing to the direction of the discipline in the field.
This study found that there had been a significant increase in interest in the field of PSD in recent years, and more papers were expected to be published in the coming years. The results showed that China and the United States had the most studies on PSD. Meanwhile, the Journal of Stroke and Cerebrovascular Diseases published the most PSD related papers and was the most important journal in the field of post-stroke depression. The results of keyword co-occurrence analysis showed that the high-frequency keywords were divided into four clusters: cluster 1(study on risk factors of PSD), Cluster 2(study on clinical correlation of PSD), cluster 3(study on mechanism of PSD) and cluster 4(treatment of PSD). Combined with the temporal evolution analysis of key words, it was revealed that PSD has shifted from the early mechanism research to the later clinical treatment and prevention research.
Funding
This work was supported by the National Natural Science Foundation of China (81874453), the Natural Science Foundation of Guangxi Province, China (2020GXNSFAA297270).The funding source had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Authors’Ethical Approval
This paper has been approved by the ethical review board of the Guangxi University of Chinese Medicine.
Contributions
W.M. ,L.X., Y.Y., Y.Z,T.X, and Z.G. contributed equally to this work.W.M. ,Y.Z.,and Z.G.conceived of the study and revised the manuscript for important intellectual content. T.X.,L.X.,B.L.,and J.L.performed the literature search and contributed all the figures. Y.Y.,X.M.,and W.M. edited the manuscript.
Informedconsent
All authors have read and approved the content of the manuscript.
Ethics approval and consent to participate Not applicable.
Consent for publication Not applicable.
Competing interests The authors declare no competing interests.