Applications of Traditional Herbal Ingredients in Skincare: Mapping the Research Landscape and Innovation Trajectories Over Four Decades

Abstract

Objective: Along with economic and technological advancements, traditional herbal ingredients have garnered increasing attention in skincare due to their antioxidative properties, low toxicity, multitargeted effects, and broad cultural acceptance. However, there remains a lack of comprehensive knowledge frameworks and systematic research trend analyses in this field. This study examines the scientific landscape and innovation patterns of herbal skincare, aiming to establish a framework for integrating traditional wisdom with modern dermatological science.

Methods: A systematic literature search was conducted in the Web of Science Core Collection (SCI-E). CiteSpace, VOSviewer, and ArcGIS were employed to analyze leading contributors, research trends, and thematic evolutions.

Results: The analysis yielded 1593 papers authored by 753 researchers from 507 institutions across 102 countries. China emerged as the top contributor, while Industrial Crops and Products was the most highly cited journal, and Zengin, Gokhan, was the most prolific author. Geographical clustering and cultural convergence characterized the distribution of countries, institutions, and authors. Keyword clustering identified 20 research networks, with recent clusters including #6 "molecular docking," #7 "chromatography-mass spectrometry," and #15 "mixture design." Persistently trending keywords included "Mechanism," "Gut microbiota," "Network pharmacology," and "Association."

Conclusion: Research on herbal skincare has evolved from raw plant usage to advanced refinement techniques, spanning animal, cellular, and molecular studies, including high-throughput screening approaches. Formulations have progressed from complex mixtures to single-compound extractions and bioactive monomers. A comprehensive research system has been established, yet standardization in dosage, application, and manufacturing remains crucial. Future studies should prioritize multicenter clinical trials and evidence-based validation. This study highlights the transition from empirical herbal applications to precision skincare, advocating for global collaboration between traditional medicine and modern regulatory science.

Keywords: healthy skin; herbs; plants; skin care.

FIGURE 1

A search strategy was used for bibliometric analysis by PRISMA.

FIGURE 2

Annual publication trends in herbal skincare research (1981–2024).

FIGURE 3

Country‐level publication analysis. (a) International collaboration network. The network spans 1981–2024 (1‐year intervals). Nodes denote countries, scaled by publication counts. Purple outlines highlight nations with intermediary centrality > 0.1, reflecting their brokerage role in cross‐border collaborations. (b) Geospatial publication density, geographic distribution mapped via ArcGIS, with node color intensity proportional to national publication output (darker hues = higher productivity).

FIGURE 4

Institutional co‐occurrence map. Nodes represent academic institutions, with size proportional to publication output. Purple‐ringed nodes denote entities exhibiting high brokerage potential (betweenness centrality > 0.1) within the cooperative landscape.

FIGURE 5

Author‐centric network mapping in CiteSpace constructed co‐authorship relationships, where node size is scaled by individual publication output.

FIGURE 6

Cumulative graph of the number of published papers in the first 10 years of the journal. The vertical axis represents the cumulative number of papers published in the journal, and the horizontal axis represents the publication time. The redder the color and the larger the node, the greater the cumulative number of papers published.

FIGURE 7

Keywords cluster analysis selected the “cluster” option and used the pathfinder algorithm to draw the connection line, so as to ensure the rationality of the cluster classification. A total of 20 keyword clusters were identified, with each cluster assigned a different color according to the timestamp in the bottom left corner. The names of the clusters were derived from a set of representative keywords with the help of the LLR algorithm.