Visualization Analysis of Research Trends and Hotspots in Inspiratory Muscle Training

Background

Inspiratory muscle training (IMT) has been studied for more than 40 years. In 1976, Leith and Bradley demonstrated that proper IMT can improve respiratory muscle strength and endurance in young people [1]. The goal of IMT is to train the inspiratory muscles based mainly on the diaphragm by applying a load resistance during the inspiratory process [2,3]. On the one hand, IMT strengthens airway clearance and defense and reduces retained secretions and collapsed airways by increasing inspiratory muscle strength and endurance, in turn reducing the occurrence of pulmonary complications and improving cardiopulmonary function [4–6]. On the other hand, IMT prevents and slows atrophy and functional decline of inspiratory muscles and avoids the blood-flow competition between respiratory and locomotor muscles during exercise by increasing the number of mitochondria and type II muscle fibers in the inspiratory muscles [7,8]. IMT also reduces activation of sympathetic nerves, improves perfusion of skeletal muscles, and increases motor function, thus promoting functional recovery of patients [9,10].

Safe, flexible, and accessible IMT is currently recommended for individual patient rehabilitation programs and is applied to relevant populations either alone or in combination with other rehabilitation modalities [11,12]. In the elderly, IMT can improve maximal inspiratory pressure (MIP) and vagal-mediated heart rate variability (HRV) [13]. Athletes can increase exercise endurance and improve training performance through IMT [14]. Clinically, IMT is mainly applied in intensive care patients, chronic obstructive pulmonary disease (COPD), stroke, heart failure (HF), spinal cord injury, and in patients who had thoracic and/or abdominal surgery [15–21]. Researchers have thus mostly focused on the rehabilitative effects, mechanism, and research overview of IMT in the above populations. They have also published relevant articles in academic journals, such as clinical trials, reviews, and meta-analyses. There is a lack of integration and systematic evaluation of these articles. IMT has recently received increased attention as an important component in respiratory rehabilitation/pulmonary rehabilitation [22,23]. Therefore, it is necessary to publish an article that comprehensively analyzes relevant articles of IMT and reflects the IMT development trends and research hotspots, which can provide scholars with important information about IMT.

Bibliometric analysis is a quantitative statistical method that uses computer image processing techniques to convert abstract data into intuitive graphics. This allows deep exploration of literature characteristics from the level of big data and can reveal the development trends and research hotspots of a certain discipline or field [24]. Recently, some researchers have used CiteSpace for bibliometric analysis. Based on analyses of collaborative networks, co-citation, and co-occurrence, CiteSpace presents the relationships among countries, institutions, journals, authors, references, and keywords through visual maps [25,26]. Scholars have performed bibliometric analysis of cardiac rehabilitation (CR) [27] and high-intensity interval training (HIIT) [28] using CiteSpace, but there are no bibliometric analyses of IMT.

Based on the Web of Science Core Collection (WoSCC) database, this study used CiteSpace software to conduct a bibliometric analysis of the literature related to IMT research between 2009 and 2022. This study provides important information for researchers related to IMT research on countries, institutions, journals, references, authors, and keywords. The information can help researchers understand hot topics in the field and explore new research directions to promote use of IMT in clinical practice, as well as cooperation and development of IMT-related disciplines.

Material and Methods

DATA SOURCE AND SEARCH STRATEGY:

The publications used in this study were retrieved from the WoSCC database, the most comprehensive database of scientific publications on global academic resources [29]. Due to daily database updates, to avoid bias the literature retrieval from WoSCC was only performed on April 3, 2023. The search terms were as follows: Topic Search=(“inspiratory muscle training”) and Time Span=2009–2022. A total of 822 publications were identified.

INCLUSION AND EXCLUSION CRITERIA:

Inclusion criteria were: (1) language of setting: English; and (2) type of paper: articles and review articles. Exclusion criteria were: (1) duplicate publications; (2) meeting abstract, letter, editorial material, correction, or book chapters; (3) literature with incomplete information; and (4) research content unrelated to IMT. A total of 558 publications were obtained using these criteria and the retrieved publications were imported into EndNote20 document management software. Two researchers read the titles and abstracts independently, excluded irrelevant publications, and cross-checked them. A third researcher assisted in the judgment when the other 2 researchers disagreed on a source. Ultimately, 504 publications (396 articles and 108 reviews) met the inclusion criteria (Figure 1), which were selected as “full record and cited references”, exported in “plain text file format”, and renamed as “download_.txt” to ensure that the CiteSpace software could correctly read these records [25].

DATA ANALYSIS:

Knowledge mapping visualization analysis was performed using CiteSpace (6.2.R2), a Java-based application developed to analyze research trends, frontiers, and hotspots in an academic field and to demonstrate the structure and distribution of scientific knowledge [30,31]. In this study, CiteSpace software was used to perform collaborative network analysis (countries/institutions/authors), co-occurrence analysis (keywords), and co-citation analysis (journals/references/authors). Results were presented as visual graphs or tables.

Nodes in the plots generated by CiteSpace represent the analyzed objects; the size of a node is positively correlated with the frequency of the object’s occurrence [30]. The internal color of a node indicates the year of the object’s occurrence and a change in color represents a change in the duration of the study [30]. Lines between 2 nodes represent the co-occurrence or co-citation relationship and the thickness of a line represents the strength of the relationship [26]. Betweenness centrality (BC) is a measure of the importance of a node in the network; a node with BC ≥0.1 is displayed with a purple ring [32]. When performing cluster analysis, Modularity value should be >0.3, otherwise the cluster is meaningless; if the Mean Silhouette (the contour value), which indicates the homogeneity of the clustering profile, is closer to 1, it means the clustering is more efficient [25]. Burst analysis represents the frequency of a node suddenly increasing compared with other nodes at a certain time in the chosen field, indicating that this node is active at a certain time, which can help determine research trends, frontiers, and hotspots [33].

Parameters of CiteSpace were set as follows: Time Slicing (2009–2022); Year per Slice (1); Term Source (Title/Abstract/Author keywords/Keywords Plus); Node Type (Author/Institution/Country/Keyword/Reference/Cited Author/Cited Journal); Top N (50); Pruning (Pathfinder/Pruning sliced networks); and Visualization (Cluster View-Static/Show Merged Network).

Results

ANNUAL QUANTITATIVE DISTRIBUTION OF PUBLICATIONS:

The annual number of publications reveals the rate of development of knowledge and research trends in the chosen field. As shown in Figure 2, the number of publications related to IMT fluctuated between 2009 and 2022, but generally showed an upward trend. Significant declines were observed from 2016 to 2017 and from 2020 to 2021, while rapid increases were observed from 2017 to 2018 and from 2019 to 2020.

COUNTRIES:

Researchers from 48 countries published articles related to IMT. The top 5 countries were Brazil (118), England (66), USA (50), China (48), and Spain (41) (Table 1). It is notable that the total number of publications from Brazil was much larger, but England had the highest centrality (BC=0.43), far exceeding that of Brazil (BC=0.05). Australia, Canada, Belgium, and Egypt were not in the top 5 for the number of publications, but they ranked high for BC value (Figure 3).

INSTITUTIONS:

A total of 295 institutions contributed IMT studies. As shown in Table 2, KU Leuven had the most publications (19). Half of the top 10 institutions were from Brazil. In terms of centrality, KU Leuven (BC=0.16) ranked first, while University Hospital Leuven (BC=0.15) and Brunel University (BC=0.13) ranked second and third, respectively (Figure 4).

JOURNALS:

As shown in Table 3, the American Journal of Respiratory and Critical Care Medicine was co-cited most frequently (336) among the top 10 journals. Most of the journals were from the United States and England, and the top 2 journals had impact factors over 30 points. Moreover, 90% of the top 10 journals were JCR Q1 or Q2 (impact factors and partitions from Journal Citation Reports – Clarivate). However, the BC values of the top 10 journals were all less than 0.1.

REFERENCES:

Reference co-citation analysis is usually used to explore research frontiers and priorities of the chosen academic field. The CiteSpace software g-index parameter k was set to 20. As shown in Table 4, half of the top 10 articles were from JCR Q1 and the top-ranked article had an impact factor of 33.801 points. The top 10 most frequently co-cited publications were clinical trials (5) or meta-analyses (5) and half of them studied the application of IMT in COPD patients.

COLLABORATIVE NETWORK ANALYSIS: A total of 393 scholars have been involved in IMT-related research. As shown in Table 5, the top 3 authors ranked by number of publications were Gosselink (9), McConnell (7), and Brown (6) and Dallago (6). The top 9 authors were from Europe (7), Oceania (1), and South America (1). In terms of centrality, all of these authors had scores less than 0.1.

CO-CITATION ANALYSIS: When 2 authors’ articles are simultaneously cited by a third author’s article, it identifies a co-citation relationship between these 2 authors [34]. Table 6 shows the top 8 authors in terms of co-citation frequency, with the first ranked Aldrich being cited 147 times. Notably, Gosselink and McConnell were ranked in the top 5 in both the co-citation analysis and the collaborative network analysis. In terms of centrality, Hill (BC=0.14) and Dallago (BC=0.10) were relatively ahead (Figure 5).

KEYWORD CO-OCCURRENCE ANALYSIS: A total of 343 keywords appeared in the articles. As shown in Table 7, the highest frequencies of co-occurrence of “inspiratory muscle training” (152) and “exercise” (122) were found more than 100 times. The centrality of key nodes was generally considered to be ≥0.1, thus the key nodes in this study included “inspiratory muscle training”, “respiratory muscles”, “breathing exercises”, and “performance” (Figure 6).

KEYWORD CLUSTERING ANALYSIS: To further determine relationships among keywords, we used the log-likelihood rate (LLR) algorithm model to perform a cluster analysis. Timeline graphs were drawn as in Figure 7 and clustering data were derived as in Table 8. IMT research areas in the study were divided into 9 clusters. The top 3 clusters were “postoperative pulmonary complications”, “exercise capacity”, and “running performance”.

KEYWORD BURST ANALYSIS: The keyword burst analysis indicates that keywords appear suddenly with higher frequencies at a certain period in this domain, and red areas indicate the length of time that a keyword has lasted since its sudden appearance [33]. The burst analysis yielded 25 keywords (Figure 8). Keywords with higher burst values were “time trial performance” (3.9), “healthy” (3.7), “weakness” (3.24), and “exercise performance” (3.19). “Diaphragm” had a maximum duration of 9 years. By the end of 2022, keywords that were still in sudden outbreak were “guidelines”, “lung function”, “surgery”, and “diagnosis”.

Discussion

STRENGTHS AND LIMITATIONS:

This study is the first bibliometric analysis of publications on IMT over the last 14 years that we are aware of. However, this study has some limitations due to the CiteSpace software. First, this study only analyzed literature related to IMT obtained from the Web of Science database between 2009 and 2022, so the inclusion of literature may not be comprehensive. Second, this study limited the language to English, and thus may have missed relevant publications in other languages. However, the WoSCC database is the most comprehensive database of scientific publications on global academic resources, “articles or reviews” are the main types of publications, and “English” is the most common language for publishing academic literatures in the worldwide. Thus, our study may address the bias caused by the above deviations.

Conclusions

This study used CiteSpace software to analyze IMT research trends and status by performing collaborative network, co-occurrence, and co-citation analyses on 504 IMT-related articles published between 2009 and 2022. The popularity of IMT research has continued to rise. COPD, inspiratory muscle, and effectiveness, have been, and continue to be, primary research topics in IMT. Currently, IMT is receiving attention in terms of the diagnosis and treatment of some diseases, lung function, surgical population, and guidelines. Researchers can further explore the optimal amount of IMT intervention, the timing of application, and the specific population effect in future studies. In addition, clinicians need to focus on the problem of IMW and make timely diagnoses so as to perform targeted IMT. In the future, IMT researchers can actively carry out multicenter and large-sample size studies to provide more high-quality, evidence-based medical findings for the field of IMT.