We carried out a bibliometric analysis using the Science Citation Index Expanded (SCIE). The search terms and strategy were as follows: (TS = (“atopic dermatitis”) OR TS = (“atopic eczema”)) AND (TS = (itch) OR TS = (Pruritus)) AND (Language = (English)), within a limited time frame, from 2014 to 2023. Among the publications, only articles and review articles were considered. The details are provided in Supplementary Table 1.

We extracted data such as titles, keywords, nations, authors, institutions, journals, and the number of citations from the screened publications. Subsequently, the data was converted to TXT format and imported into the data into the Online Analysis Platform of Bibliometry,¹ VOSviewer v1.6.10.0, and CiteSpace (version 5.8. R3) for further analysis.

We performed a bibliometric analysis of various literature characteristics, including countries, institutions, authors, journals, references, and keywords. The number of publications (NP) was used as a measure of productivity, while the number of citations (NC) excluding self-citations served as an indicator of impact. The average citations per year (ACY) score was adopted to eliminate time bias. The H-index was used to determine a researcher’s academic contribution (2). The Impact Factor (IF) is a widely acknowledged measure for evaluating the quality and impact of medical journals (3). The Global Citation Score (GCS) is a crucial measure of an article’s contribution (4). We generated bibliometric maps using VOSviewer to obtain comprehensive information through the analysis of co-citation and co-occurrence (5). In CiteSpace, we employed cluster analysis, timeline, references, and keyword bursts to visually assess the current situation and identify emerging trends in the domain (4).

In total, 2,534 articles published from 2014 to 2023 were retrieved from the Science Citation Index Expanded (SCI-E) of the Web of Science Core Collection (WoSCC). Figure 1 shows the geographical distribution of publications on AD-related itch, with most contributions originating from countries in Europe, America, and Asia. Figure 2 shows the number of annual publications, which increased from 130 in 2014 to 402 in 2023. A polynomial-fitting curve depicts the yearly trend in publication volume. In general, the increase in the NP indicates that AD-related itch research has experienced rapid development recently. The top 10 countries were ranked by NP and are presented in Supplementary Table 2. The country that published the majority of papers was the United States, followed by Germany, Japan, China, and South Korea. The United States had the highest number of publications, with 924 publications, a total of 29,248 citations, and the highest H-index (88). Germany had the highest ACY (48.6). In Figure 3, the size and hues of the circle symbolize the volume of the papers. The larger the volume of papers published, the larger the circle and the hues shift from light blue to red. The NP in China showed a significant increase, and the United States contributed the most to this research field.

Supplementary Table 3 presents the top 10 most productive affiliations. The top seven affiliations were all from the United States. Among them, Northwestern University had the highest NP (148). The Feinberg School of Medicine had the highest NC (16296), H-index (50), and ACY (275.37).

Supplementary Table 4 lists the top 10 most productive authors. They were primarily from the United States, Germany, Japan, and Denmark. They authored 541 publications, which constituted 21.3% of the overall number of papers. Simpson, EL from Oregon Health & Science University had the highest NP (92), NC (6738), and H-index (37).

We identified the top 10 journals with the highest number of publications, as listed in Supplementary Table 5. Acta Dermato-Venereologica had the highest NP (93). The Journal of Allergy and Clinical Immunology recorded the highest NC (3975), and the Journal of the American Academy of Dermatology had the highest H-index (32).

Figure 4 shows the GCS’s annual number for the top 10 most cited articles. The size and hues of the circle indicate the GCS of the papers. The article titled Unbiased Classification of Sensory Neuron Types by Large-scale single-cell RNA Sequencing was cited the most, followed by Atopic Dermatitis and Dupilumab Treatment in Adults with Moderate-to-Severe Atopic Dermatitis. The third article had a marked increase in GCS, especially during 2022–2023. This finding indicates the significant influence this article has had in recent years.

Figure 5 shows the top 10 productive categories of AD-related itch. Dermatology emerged as the most prevalent research area with 1,160 papers, followed by immunology (333 papers), allergy (260 papers), and pharmacology (245 papers). The aforementioned research themes have been identified as research trends over the past decade.

The networks of co-citation emphasize research topics in this specific field. We analyzed 2,534 publications and 59,994 references with CiteSpace software. We conducted a clustering analysis to ensure the homogeneity of highly cited literature on AD-related itch. Figure 6A provides a map of co-citation networks. Each node symbolizes a single article, with its size directly proportional to the frequency of citation. The connection between nodes indicates that they were cited in the same literature. The thickness of the lines indicates the degree of relevance between citations. The shorter the line, the stronger the association between the two citations. In addition, nodes with different colors were used to segregate the articles into distinct clusters. Of the 59,994 references, 154 (classified into 4 clusters) had been cited at least 50 times. Cluster 1 (in red) references primarily encompass the pathogenesis, epidemiology, clinical manifestations, methods for diagnosis and assessment, and treatment methods of AD (6). Cluster 2 (in green) references predominantly address the mechanism of AD, introducing novel mechanisms such as type 2 cytokines (7), interleukin (IL)-31, transient receptor potential vanilloid 1 (TRPV1), transient receptor potential cation channel subfamily A member 1 (TRPA1), and sensory neurons (8, 9). Cluster 3 (in blue) references addressed that dupilumab could significantly improve symptoms of AD (10) and quality of life (11). Cluster 4 (in yellow) focused on the diagnostic criteria that could be applied to dogs with AD (12). Figure 6B shows articles with a high frequency of co-citation in a map of density visualization. Articles in yellow areas are cited more frequently, whereas those in green are cited less frequently. The articles that were primarily published were collected from the following journals: The Lancet, New England Journal of Medicine, Journal of the American Academy of Dermatology Allergy, and Journal of Clinical Investigation. A timeline view of co-citation analysis for AD-related itch is shown in Figure 6C. The figure indicates the top 10 clusters. Each horizontal line represents a cluster, with the connections between the nodes denoting co-citation relationships. Timeline, positioned at the top, progresses from left to right, indicating the chronological development from earlier to more recent publications. The cluster label on the right presents the topic of the domain. The 10 clusters included #0crisaborole, #1 prurigo nodularis, #2 itch, #3 upadacitinib, #4 severity, #5 skin barrier, #6 interleukin (IL)-31, #7 dupilumab, #8 dog, and #9 stress. The citation burst of reference refers to an article exhibiting a sudden surge in citation frequency, which generally signals the emergence or a shift in the field. Furthermore, the reference with higher burst strength indicates greater significance within the field (13). Figure 6D displays the top 20 strongest citation bursts of references in AD-related itch research. The blue bars depict the publication periods of the references, while the red bars illustrate the periods of citation burst.

Among the burst references, the latest article was named Anti-IL-31 Receptor A Antibody for Atopic Dermatitis, published in The New England Journal of Medicine in 2017, and the burst occurred from 2018 to 2019. IL-31 secreted by T-helper type 2 cells is an influential pruritogenic cytokine (14). During the phase 2 trial, patients with moderate-to-severe AD who received monthly treatment with nemolizumab showed marked alleviation in the symptom of pruritus, thus demonstrating that nemolizumab effectively targets IL-31 receptor A in AD (15). The publication with the highest burst strength, which was published in the New England Journal of Medicine, reported that dupilumab could dramatically treat patients with AD without notable adverse reactions (5).

We analyzed 7,078 keywords collected from 2,534 publications. In Figure 7A, the node symbolizes the keyword, and the node’s size reflects the frequency of keyword occurrence. The lines connecting the nodes signify the co-occurrence of keywords. The closer the relationship between nodes, the thicker the line. There are 156 words that appeared at least 25 times and were categorized into 4 distinct clusters using VOSviewer. Cluster 1 (in red) is primarily concerned with the mechanism of AD-related itch. Cluster 2 (in green) focused on the epidemiology, burden, symptoms, and guidelines of eczema. Cluster 3 (in blue) is related to the complications of AD, including food allergies, asthma, and rhinitis. Cluster 4 (in yellow) focused on dupilumab, nemolizumab, and other biologics in AD treatment. Figure 7B maps the overlay of keywords. The varying colors represent the respective publication years, and keywords in yellow emerged later than those in purple. Figure 7C is a clustering timeline of the co-occurrence of keywords. The figure presents the top eight clusters. The connections between the nodes indicate the co-occurrence of keywords. The eight clusters included #0 quality of life, #1 sensory neuron, #2 dupilumab, #3 dog, #4 expression, #5 uremic pruritus, #6 contagious itch, and #7 skin barrier. Even across species, the incidence of AD is gradually increasing. Considerable research has been conducted on the prevalence, impact, and treatment of canine AD (16). Considering the similarities in mechanism between pruritus in AD and uremic pruritus, studies on uremic pruritus are gradually emerging (17). The analysis of keyword bursts as a valuable method could help discern trends in a field. Figure 7D shows the top 20 strongest keyword bursts on AD-related itch over the past decade. The keyword “Nc/Nga mice” had the strongest burst (7.79). The Nc/Nga mice are a suitable animal model, as they display clinical and histological similarities to human AD. The latest keywords were “protein-coupled receptor” and “Janus kinases (JAK) inhibitor” in 2020. “Protein-coupled receptor,” “jak inhibitor,” “in vitro,” and “adult patient” were hot keywords that appeared in recent years and could help scholars understand research hotspots. The map of density visualization for keywords co-occurrence showed that, except for the terms “atopic dermatitis” and “itch,” other high-frequency keywords included “expression,” “efficacy,” “dupilumab,” “double-blind,” “skin,” “inflammation,” and “children.”